Micro-particle three-dimensional positioning method, system, device and medium based on local reconstruction

Abstract

The application discloses a particle three-dimensional positioning method, system and device based on local reconstruction and a medium, and belongs to the field of three-dimensional imaging measurement. The method comprises the following steps: acquiring real-time interference images containing a plurality of particle samples; enhancing two-dimensional features by using reconstruction and superposition of a plurality of layers of plane light intensity, and performing two-dimensional positioning of the samples; performing local interference graph cutting and three-dimensional reconstruction according to the two-dimensional positioning to calculate axial features, and then acquiring longitudinal coordinates of corresponding positions; and obtaining continuous three-dimensional trajectories according to three-dimensional coordinates in continuously obtained interference images based on space-time continuity. The application uses an interference optical imaging method to record three-dimensional information in a field of view, can realize three-dimensional continuous tracking of a plurality of samples, has the characteristics of fast calculation, high positioning accuracy and good robustness, and is favorable for practical application and popularization of related technologies.

Description

Technical Field

[0001] This invention relates to the field of imaging measurement, and in particular to a method, system, device, and medium for rapid three-dimensional localization of microparticles based on local reconstruction. Background Technology

[0002] Particulate samples are widely distributed in the natural environment and daily life and production, and their location tracking is closely related to many core issues such as materials characterization, flow field research, and life sciences. As research deepens, related studies have shifted from two-dimensional observation to three-dimensional tracking in the real world. However, the lack of observation instruments limits research in this field.

[0003] Interferometric imaging, as an optical characterization method, is well-suited for tracking and studying particulate samples due to its high precision, lack of marker requirements, and single-frame positioning capabilities. This type of method preserves the light field fluctuations along the propagation direction by retaining phase information in the form of interference fringes, enabling subsequent calculations to reconstruct the sample's height information. To accelerate computation, various positioning algorithms have been proposed. For example, Gong Xiangjun et al., in their Chinese invention patent CN114926529A, optimized the algorithm to avoid the large amount of computation caused by repeated comparisons. However, these methods all require overall reconstruction analysis, necessitating calculations and comparisons of a large amount of background information unrelated to the observed sample. The sheer volume of data to be processed places high demands on computing power, especially during batch experiments or when using high-speed cameras for recording. The high time cost significantly limits the rapid progress of research. Furthermore, this patent only optimizes subsequent positioning calculations and cannot reduce the most time-consuming aspect: overall three-dimensional light intensity reconstruction.

[0004] It is evident that optimizing computational methods to obtain a high-precision and fast 3D positioning method is of great significance for achieving rapid positioning calculations of particle interferometric images. This rapid 3D positioning method can further expand its application in related research and production inspection scenarios. Summary of the Invention

[0005] To at least partially address one of the technical problems existing in the prior art, this invention provides a method, system, device, and medium for rapid three-dimensional localization of particles based on local reconstruction. By first performing two-dimensional local positioning to define the position and then replacing full reconstruction with multiple small-area local reconstructions, the calculation speed can be accelerated.

[0006] To achieve the objective of this invention, a rapid three-dimensional localization method for particles based on local reconstruction is provided, comprising the following steps:

[0007] Interferometric optical imaging methods are used to record interferometric images of samples containing multiple particles in the field of view;

[0008] The height range of the particle sample is evaluated, and the light field is reconstructed within the height range to obtain several layers of two-dimensional reconstructed plane light fields to reflect the spatial light field distribution within the height range. The difference in reconstruction distance between adjacent planes is equal and they are uniformly distributed within the observation range.

[0009] The light intensity values ​​of the reconstructed plane light field of each layer are superimposed, and the two-dimensional projection coordinates of the particle sample are obtained based on the superposition result.

[0010] Centered on the two-dimensional projection coordinates of the particle sample, a local interferogram containing the main information of the particle sample is cropped from the original interferogram, and the light intensity variation curve of the two-dimensional projection coordinate position in the z direction is reconstructed based on the local interferogram.

[0011] Feature extraction is performed on the light intensity variation curve in the z-direction to obtain the z-coordinate of the two-dimensional projection coordinates. The three-dimensional coordinates of the particle sample are then determined by combining the two-dimensional projection coordinates of the particle sample.

[0012] The three-dimensional trajectories of multiple particles in the field of view are obtained based on the spatiotemporal continuity and the distribution of three-dimensional coordinate points.

[0013] Then, motion-related parameters such as speed, acceleration, and motion pattern are calculated.

[0014] Furthermore, the interferometric optical imaging method uses coherent light of known wavelength to illuminate the image via plane waves or spherical waves, thereby obtaining an interferogram that is defocused and has a defocus distance within a certain range.

[0015] Furthermore, the plane light field and the z-axis light intensity variation curves were obtained by computer numerical reconstruction based on the optical diffraction propagation formula.

[0016] Furthermore, methods for reconstructing multi-layer planar light fields include:

[0017] Based on the actual experimental conditions, the range of the particulate sample is estimated, and the starting distance D1 and the ending distance D2 of the subsequent reconstruction calculation are determined to ensure that the z-direction position of the particulate sample in the field of view is within the reconstruction range [D1,D2] during subsequent studies.

[0018] Based on D1 and D2, generate an arithmetic sequence with D1 and D2 as the first and last elements, containing N elements (the specific value of N is specified in advance, generally not exceeding 30). Each element in this sequence is used as the reconstruction distance for the planar light field of each layer.

[0019] The planar light intensity distribution at each reconstruction distance is calculated based on the optical diffraction propagation formula, and N discrete light intensity distribution maps with the same size as the original interferogram are obtained.

[0020] Furthermore, when performing light intensity superposition, it is necessary to sum the discrete values ​​according to their positions in the image to obtain a discrete matrix with the same size as the original image. Then, a local maximum search is performed to judge all elements in the superposition matrix. When a point is larger than a certain range centered on that point, the point is determined to be a possible two-dimensional position of the particle.

[0021] Furthermore, the local interferogram extraction and 3D reconstruction include:

[0022] The cropping template size is set according to the defocused image of the particles to ensure that the main interference signal of the particles in the center of the image can be preserved when extracting the local interferogram in the subsequent process.

[0023] Centered on the two-dimensional position of each particle, a corresponding number of local interferograms are obtained independently according to the cutting template. When the position to be obtained exceeds the original interferogram due to the position of the two-dimensional positioning point being close to the edge, the excess part is filled with zero value, image average value, etc. as appropriate.

[0024] Numerical reconstruction is performed on each local interferogram according to the optical diffraction propagation formula, and the light intensity change curve of its center position (i.e. the two-dimensional position of the particle) is extracted as it propagates along the z direction. The reconstruction distance difference between adjacent points of the curve is consistent with the side length of the pixel unit. The reconstruction distance range is [D1, D2] to ensure that the curve contains the change process of particle defocusing-focusing-defocusing.

[0025] Furthermore, when performing three-dimensional positioning, the z-coordinate can be obtained through various methods, including but not limited to local peak finding of light intensity and curve fitting.

[0026] Furthermore, the method for obtaining the three-dimensional trajectory includes the following steps:

[0027] Set the maximum allowable distance for trajectory connection. If the distance between the positioning coordinates on adjacent frames is less than this value, the two points can be considered to be the same sample.

[0028] Arrange all 3D positioning results in chronological order. Using the positioning results of the first frame as the center, search for 3D positioning results in the second frame that satisfy the condition that the distance is less than a preset number. After judging all results, use all positioning points in the second frame as the starting point to search for corresponding points in the next frame. Repeat this process until all continuously recorded interferometric images are judged.

[0029] The trajectory is connected based on the relationship between the points. Points that meet the judgment conditions are arranged into a three-dimensional motion trajectory in time sequence. When multiple points are connected to the same point or one point is connected to multiple points, a second judgment is required to determine the true trajectory. At the same time, trajectories that are too short should be discarded to avoid introducing erroneous results.

[0030] Furthermore, when performing three-dimensional light field reconstruction, the relevant convolution factors can be calculated in advance and the calculation accuracy can be appropriately reduced to further accelerate the calculation speed.

[0031] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0032] (1) The present invention adopts a step-by-step local reconstruction method to achieve rapid three-dimensional positioning and tracking of particles. Two-dimensional positioning is performed first, and then the z-direction features at the local position are calculated. This avoids the reconstruction of the three-dimensional light field of a large number of blank background parts, thereby greatly improving the calculation speed of the three-dimensional positioning and tracking process of particle interference images.

[0033] (2) The present invention can realize rapid three-dimensional positioning and tracking calculation of various particle samples. Since it does not require calculation and analysis of a large proportion of blank background in the field of view, compared with the global reconstruction method, it can reduce the calculation time to less than one-tenth of the global reconstruction method while maintaining the positioning accuracy. It also greatly reduces the memory and video memory required for calculation, reduces time and hardware costs, and makes it possible to perform data processing and three-dimensional observation in real time on lighter and cheaper devices.

[0034] (3) The method of the present invention uses only the local interferogram near the sample for reconstruction during positioning, which can eliminate the interference of high-intensity signals at a distance. It plays a certain role in improving the accuracy of the calculation results of low signal-to-noise ratio interferometric images and has high reference value for solving special problems. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following description includes accompanying drawings of the relevant technical solutions in the embodiments of the present invention or the prior art. It should be noted that the accompanying drawings described below are merely for the purpose of clearly illustrating some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0036] Figure 1 This is a flowchart of a rapid three-dimensional localization method for microparticles based on local reconstruction in this invention;

[0037] Figure 2 This is a schematic diagram of the defocused interference image of the plastic microspheres obtained by the interferometric optical imaging method in Embodiment 1 of the present invention;

[0038] Figure 3 This is a schematic diagram of the three-dimensional positioning results of multiple microspheres in a single interferogram in Embodiment 1 of the present invention;

[0039] Figure 4 This is a schematic diagram of the defocused interference image of Escherichia coli in Embodiment 2 of the present invention;

[0040] Figure 5 This is a schematic diagram of a three-dimensional trajectory image of Escherichia coli obtained by calculation in Embodiment 2 of the present invention. Detailed Implementation

[0041] The embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. The step numbers in the following embodiments are set only for ease of explanation, and there is no limitation on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0042] Please see Figure 1 The present invention provides a method for rapid three-dimensional localization of particles based on local reconstruction, comprising the following steps:

[0043] Step S1: Use interferometric optical imaging to record an interferometric image of a sample containing multiple microparticles in the field of view.

[0044] In this step, the optical interferometric imaging method uses coherent light of known wavelength for illumination via plane waves or spherical waves to obtain an interferometric image that is defocused within a certain range. This allows the sample in the obtained image to exhibit interference ring characteristics and also enables good reconstruction of the sample morphology through simulation. The defocus distance is determined by the specific imaging system and, in some optional embodiments, is between 5 and 20 micrometers.

[0045] Step S2: Evaluate the height range of the particle sample and reconstruct several planar light fields within the height range to reflect the spatial light field distribution within that height range, wherein the difference in reconstruction distance between adjacent planes is equal and they are uniformly distributed within the observation range.

[0046] In this step, the planar light field and the z-axis light intensity variation curve of step S4 are both obtained by numerical reconstruction by computer based on the optical diffraction propagation formula.

[0047] In this step, the method for reconstructing the multi-layer planar light field includes the following steps:

[0048] Step S2.1: Estimate the range of the particulate sample based on the actual experimental conditions, and determine the starting distance D1 and ending distance D2 for subsequent reconstruction calculations to ensure that the z-axis position of the particulate sample in the field of view is within the reconstruction range [D1,D2] during subsequent studies.

[0049] Step S2.2: Generate an arithmetic sequence containing N elements, with the starting distance D1 and the ending distance D2 as the beginning and end. Use each element in the arithmetic sequence as the reconstruction distance of the planar light field of each layer. The specific value of N is specified in advance. In some optional embodiments, N does not exceed 30.

[0050] Step S2.3: Calculate the planar light intensity distribution at each reconstruction distance according to the optical diffraction propagation formula to obtain N layers of discrete light intensity distribution maps with the same size as the original interferogram.

[0051] Step S3: Superimpose (sum) the intensity values ​​of the reconstructed planar light fields of each layer to obtain the superposition result, and obtain the two-dimensional projection coordinates of the particle sample by searching for the local maximum value of the superposition result.

[0052] In this step, the light intensity superposition requires summing the corresponding light intensity values ​​at their respective positions on the reconstruction plane, resulting in a discrete matrix with the same size as the original image. Then, a local maximum search is performed, evaluating all elements in the superposition matrix. When a point is larger than a certain range centered on that point, it is determined to be a possible two-dimensional location of the particle, yielding the two-dimensional projected coordinates of the particle sample. Edge locations need to be excluded in this calculation to avoid misclassification. In some optional embodiments, this certain range is determined by the sample size and concentration, and needs to be determined based on experimental conditions.

[0053] Step S4: Using the two-dimensional projection coordinates obtained in step S3 as the center, cut out a local interferogram containing the main information of the particle sample from the original interferogram, and reconstruct the light intensity variation curve of the center position of the local interferogram in the z direction based on the local interferogram.

[0054] In this step, the extraction and 3D reconstruction of the local interferogram includes the following steps:

[0055] Step S4.1: Set the cropping template size according to the defocused particle image in step S1 to ensure that the obtained local interferogram can retain the main interference signal of the particle sample in the center of the image when extracting the local interferogram later.

[0056] Step S4.2: Using the sample two-dimensional position of each particle obtained in step S3 as the center, obtain the corresponding number of local interferograms independently according to the cutting template. In some optional embodiments, when the position to be obtained exceeds the original interferogram due to the position of the two-dimensional positioning point being close to the edge, the excess part is filled in the form of zero value, image average value, etc., as appropriate.

[0057] Step S4.3: Perform numerical reconstruction on each local interferogram according to the optical diffraction propagation formula, and extract the light intensity change curve of the center position (i.e., the two-dimensional position of the particle sample) of each local interferogram when it propagates along the z direction. The reconstruction distance difference between adjacent points of the light intensity change curve is consistent with the side length corresponding to the pixel unit. The reconstruction distance range is [D1, D2] to ensure that the light intensity change curve contains the change process of particle defocusing-focusing-defocusing.

[0058] Step S5: Extract features from the light intensity change curve in the z-direction to obtain the z-coordinate of the two-dimensional projection coordinates. Combine the two-dimensional positioning results obtained in step S3 to determine the three-dimensional coordinates of the particle sample.

[0059] In this step of 3D positioning, the z-coordinate can be obtained through various methods, including but not limited to local peak finding of light intensity and curve fitting.

[0060] Step S6: Based on the spatiotemporal continuity and the distribution of three-dimensional coordinate points, obtain the three-dimensional trajectories of multiple particles in the field of view, which can facilitate the calculation of motion-related parameters such as velocity, acceleration, and motion mode.

[0061] In this step, the calculation of the three-dimensional trajectory includes:

[0062] A maximum distance is allowed for trajectory connection. If the distance between the positioning coordinates on adjacent frames is less than the maximum distance, the two points can be considered to be the same sample.

[0063] Arrange all three-dimensional positioning results in chronological order. Using the positioning results of the first frame as the center, search for whether there are three-dimensional positioning results in the second frame that satisfy the condition that the distance is less than the maximum distance allowed for the trajectory connection. After judging all results, use all positioning points in the second frame as the starting point to search for corresponding points in the next frame. Repeat this process until all continuously recorded interferometric images are judged.

[0064] The trajectory is connected based on the relationship between the points. The positioning points that were regarded as the same sample in the previous step are arranged into a three-dimensional motion trajectory according to the time sequence. When multiple points are connected to the same point or one point is connected to multiple points, a second judgment is required to determine the true trajectory. At the same time, trajectories that are too short should be discarded to avoid introducing erroneous results.

[0065] In some optional embodiments, when performing the three-dimensional light field reconstruction, the relevant convolution factors can be calculated in advance and the calculation accuracy can be appropriately reduced to further accelerate the calculation speed.

[0066] This invention obtains multiple local interferograms corresponding to the number of particle samples, and correspondingly obtains the same number of light intensity change curves, thereby obtaining the same number of three-dimensional coordinates of particles. Then, the three-dimensional trajectory in step S6 can be obtained through the three-dimensional coordinates of all particles.

[0067] Example 1

[0068] This embodiment provides experimental results of three-dimensional localization of a group of polystyrene (PS) microspheres adsorbed on a glass slide surface using a rapid three-dimensional localization method for microparticles based on local reconstruction as described in this invention. The method will be explained in detail below with reference to the accompanying drawings and specific embodiments.

[0069] This invention utilizes a coaxial holographic imaging system to record defocused interference patterns of PS plastic microspheres. The light source used in the experiment was an LED light source with a wavelength λ = 505 nm, and the holographic imaging system had a magnification of 40x. The camera used for imaging was a monochrome camera with a pixel side length of 6.5 μm. The sample used was PS plastic microspheres adsorbed onto the surface of a glass slide. After manual focusing and defocusing, the obtained interference image is shown below. Figure 2 As shown.

[0070] Based on the defocus distance, this embodiment of the invention determines that the sample's position is within the range of 3.5 μm to 63.5 μm from the imaging plane, and subsequent calculations are based on this range. First, a relatively sparse overall three-dimensional light field reconstruction is performed. Empirical data shows that superimposing 16 layers of light fields can achieve relatively accurate two-dimensional positioning. The light field is then divided into equal intervals according to a preset range and reconstructed in a planar manner. Specifically, a three-dimensional discrete light field is obtained by simulating the optical propagation process using the Rayleigh-Sommerfeld diffraction formula at distances of 3.5 μm, 7.5 μm, 11.5 μm, ..., 63.5 μm.

[0071] The three-dimensional discrete light fields obtained from a single frame image are superimposed along the z-direction, that is, the obtained sets of planar light intensities are summed to obtain a superimposed image with the same pixel size as the original image. Local maximum search is performed on the entire superimposed image to find the positions where the values ​​are significantly greater than their own surroundings and the background, and these are regarded as the two-dimensional projected coordinates of the microsphere sample in the field of view.

[0072] Based on the size of the diffraction ring after propagation from the microsphere, it is assumed that a 161*161 pixel side-length field of view centered on the microsphere can completely retain the main interference information, sufficiently reflecting its height information. A 161*161 pixel size is set as the template for cutting local interferograms. Using the sample positions obtained from two-dimensional positioning as centers, local interferograms are obtained from the original interferograms according to the cutting template, with the number of images matching the two-dimensional positioning points. These local interferograms are reconstructed, and their light intensity signals in the range of 3.5μm to 63.5μm are calculated. To ensure consistent data granularity, the distance between adjacent reconstructed light intensity images matches the pixel side length and magnification factor, which in this embodiment is [3.5, 3.6625, 3.825, ..., 63.375], in μm. After obtaining the z-axis light intensity variation curve, it is fitted to obtain the peak position, yielding the z-axis coordinates of each point. The two-dimensional positioning results and their corresponding z-axis positioning results are organized to obtain the three-dimensional positioning results of the microsphere in the field of view, as shown below. Figure 3 As shown in the results, the three-dimensional positioning of the microspheres is basically at the same height, which is consistent with the facts, proving the effectiveness of the positioning method of this invention. The calculation time of a single interferogram is tens of times faster than similar methods under the same hardware conditions, enabling rapid positioning of microparticle samples.

[0073] Example 2

[0074] This embodiment provides experimental results of three-dimensional localization of a group of E. coli dilutions using a rapid three-dimensional localization method for particles based on local reconstruction as described in this invention. The method will be explained in detail below with reference to the accompanying drawings and specific embodiments.

[0075] This invention utilizes a coaxial holographic imaging system to record defocused interferograms of E. coli samples. The light source used in the experiment was an LED light source with a wavelength λ = 505 nm, and the holographic imaging system had a magnification of 40x. The camera used for imaging was a monochrome camera with a pixel side length of 6.5 μm. After manual focusing and defocusing, the original interferometric image obtained is shown below. Figure 4 As shown.

[0076] Based on the actual defocus distance, this embodiment of the invention determines that the sample's position is within the range of 3.5 μm to 63.5 μm from the imaging surface, and subsequent calculations are based on this range. First, a relatively sparse overall three-dimensional light field reconstruction is performed. Empirical data shows that superimposing 16 layers of light fields can achieve relatively accurate two-dimensional positioning. The light field is then divided into equal intervals according to a preset range and reconstructed in a planar manner. Specifically, a three-dimensional discrete light field is obtained by simulating the optical propagation process using the Rayleigh-Sommerfeld diffraction formula at distances of 3.5 μm, 7.5 μm, 11.5 μm, ..., 63.5 μm.

[0077] The three-dimensional light field obtained from a single frame image is superimposed along the propagation direction to obtain a longitudinal light intensity superimposed map with the same pixel size as the original image. Local maximum search is performed on the entire superimposed map to find points that are significantly larger than the surrounding area and background. The location of these points is regarded as the two-dimensional projection coordinates of E. coli in the field of view.

[0078] Based on the size of the diffraction ring after the spread of *E. coli*, it is considered that a field of view with a side length of 181*181 pixels centered on *E. coli* is sufficient to reflect its height information; therefore, 181*181 pixels is set as the size for cutting the local interferogram. Using the sample position obtained from the two-dimensional localization as the center, local interferograms are obtained from the original interferogram according to the set size, with the number of images consistent with the two-dimensional localization points. The light field within a preset range of 3.5μm to 63.5μm is reconstructed from the center of these interferograms. The distance between adjacent reconstructed light intensity images matches the pixel side length and magnification, which in this embodiment is [3.5, 3.6625, 3.825, ..., 63.375], in μm. The longitudinal coordinate of each sample is obtained by converting the peak position of the light intensity change curve. By organizing the two-dimensional localization results and their corresponding height localization results, the three-dimensional localization results of *E. coli* in the field of view can be obtained.

[0079] The obtained 3D positioning points are connected. Points within a 5μm distance are considered to be on the same trajectory. Using the 2D positioning point of the first frame as the center, the search is performed to see if a corresponding point exists in the next frame. If no positioning point exists within the preset distance, the trajectory search is terminated. If only one positioning point exists within the preset distance, it is considered to be on the same trajectory as the starting point. If two or more positioning points exist within the preset distance, they are initially assumed to be corresponding points. The connection method of the 3D trajectories between consecutive frames is determined based on the subsequently calculated distance between points, the spatiotemporal consistency of the motion, and the final trajectory length. This step is repeated to determine the 3D trajectory of *E. coli*. Figure 5 One of the three-dimensional trajectories is shown. It can be observed that the result achieves continuous three-dimensional tracking of moving *E. coli* bacteria in the field of view, proving the effectiveness of the method. The computation time for a single interferogram is tens of times faster than similar methods under the same hardware conditions. In the experiment, the average processing time for a single image was less than 0.2 seconds, and it is expected that real-time three-dimensional tracking can be achieved under higher hardware conditions.

[0080] In summary, the foregoing embodiments of the present invention have at least the following advantages and beneficial effects compared to the prior art:

[0081] (1) This method uses a distributed positioning layout reconstruction three-dimensional positioning method, which can quickly realize the three-dimensional positioning of discrete samples in space, especially the three-dimensional positioning of sparse particles in a large field of view, effectively saving the computational cost of related studies and maintaining high positioning accuracy.

[0082] (2) This method obtains the distribution of the sample in the planar direction by reconstructing and superimposing a small amount of planar light intensity. Its contrast is higher than that of the original interferogram and there is no path dependence, which can effectively realize the two-dimensional positioning of the sample.

[0083] (3) In the distributed positioning of this method, the determination of the planar position is completely independent, which can easily introduce various traditional positioning algorithms, so that it has the characteristic of quickly solving the problems caused by special samples in practical applications, and thus has high application value.

[0084] (4) This method obtains the z-direction light field features of the sample by local image segmentation and reconstruction, which effectively solves the problem that traditional methods require a lot of time for global reconstruction, and also effectively preserves the light intensity changes brought about by the sample in the propagation direction. Therefore, it can not only achieve localization, but also easily extract relevant features of the sample for comparative evaluation.

[0085] The rapid three-dimensional localization method for particles based on local reconstruction in the foregoing embodiments of the present invention comprises the following steps: Figure 1 As shown, any combination of steps in the executable method embodiments possesses the corresponding functions and beneficial effects of the method. In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated, wherein the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0086] Example 3

[0087] This invention provides a rapid three-dimensional localization system for particles based on local reconstruction, used to implement the method provided in the foregoing embodiments. The system includes the following modules:

[0088] The recording module is used to record interference images of multiple particle samples in the field of view using interferometric optical imaging methods;

[0089] The light field plane reconstruction module is used to evaluate the height range of the particles and perform light field reconstruction within the height range to obtain several layers of two-dimensional reconstruction plane images to reflect the spatial light field distribution within the height range. The reconstruction distance difference between adjacent planes is equal and they are uniformly distributed within the observation range.

[0090] The two-dimensional positioning module is used to superimpose the light field intensity values ​​of the reconstructed planes of each layer, and obtain the two-dimensional projection coordinates of the particles by searching for local maxima in the superposition results.

[0091] The local three-dimensional light field reconstruction module is used to crop out a local interferogram containing the main information of the particle sample from the original interferogram with the two-dimensional projection coordinates as the center, and reconstruct the light intensity change curve of the two-dimensional projection coordinate position in the z direction based on the local interferogram.

[0092] The three-dimensional positioning module is used to extract features from the light intensity change curve in the z-direction, obtain the z-coordinate of the two-dimensional projection coordinate, and determine the three-dimensional coordinate of the particle by combining the two-dimensional positioning results.

[0093] The trajectory tracking module is used to obtain the three-dimensional trajectories of multiple particles in the field of view based on the spatiotemporal continuity and the distribution of three-dimensional coordinate points.

[0094] Example 4

[0095] This invention provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method provided in the foregoing embodiments.

[0096] Example 5

[0097] This invention provides a computer-readable storage medium storing a computer program or instructions thereon, which, when executed by a processor, implement the steps of the method provided in the foregoing embodiments.

[0098] Furthermore, although the invention has been described in the context of functional modules, unless otherwise stated, one or more of the stated functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. A detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, the actual implementation of the module will be understood within the conventional skills of an engineer, taking into account the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein. Therefore, those skilled in the art can implement the invention set forth in the claims using ordinary techniques without excessive experimentation. Additionally, the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0099] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A method for rapid three-dimensional localization of particles based on local reconstruction, characterized in that, Includes the following steps: Interferometric optical imaging methods are used to record interferometric images of samples containing multiple particles in the field of view; The height range of the particle sample is evaluated, and the light field is reconstructed within the height range to obtain several layers of two-dimensional reconstructed plane light fields to reflect the spatial light field distribution within the height range. The difference in reconstruction distance between adjacent planes is equal and they are uniformly distributed within the observation range. The light intensity values ​​of the reconstructed plane light field of each layer are superimposed, and the two-dimensional projection coordinates of the particle sample are obtained based on the superposition result. Centered on the two-dimensional projected coordinates of the particle sample, a local interferogram containing the main information of the particle sample is cropped from the original interferogram, and the two-dimensional projected coordinate position is reconstructed based on the local interferogram. z Curve of light intensity variation in direction; right z Feature extraction is performed on the light intensity variation curve along the direction to obtain the two-dimensional projection coordinates of the location. z The three-dimensional coordinates of the particle sample are determined by combining the two-dimensional projection coordinates of the particle sample with the coordinates of the coordinates. The three-dimensional trajectories of multiple particles in the field of view are obtained based on the spatiotemporal continuity and the distribution of three-dimensional coordinate points.

2. The rapid three-dimensional localization method for particles based on local reconstruction according to claim 1, characterized in that, The planar light field and the z The light intensity variation curves were all obtained by computer numerical reconstruction based on the optical diffraction propagation formula.

3. The rapid three-dimensional localization method for particles based on local reconstruction according to claim 1, when performing multi-layer two-dimensional planar light field reconstruction, the determination of the reconstruction distance includes the following process: Estimate the range of the particulate sample to determine the starting distance D1 and ending distance D2 for subsequent reconstruction calculations; Based on D1 and D2, generate a set of arithmetic sequences containing N elements, with the starting distance D1 and the ending distance D2 as the beginning and end. Use each element in the arithmetic sequence as the reconstruction distance for each layer of planar light field. Calculate the planar light intensity distribution at each reconstruction distance to obtain N layers of discrete light intensity distribution maps with the same size as the original interferogram.

4. The rapid three-dimensional localization method for particles based on local reconstruction according to claim 1, characterized in that, When performing light intensity superposition, the corresponding sums are performed according to the position of the light intensity value on the reconstruction plane, and the final result is a discrete matrix with the same size as the original image.

5. The rapid three-dimensional localization method for particles based on local reconstruction according to claim 4, characterized in that, The two-dimensional projection coordinates of the particle sample are obtained by performing a local maximum search on the superposition result. The local maximum search is performed by judging all elements in the discrete matrix obtained by superposition to obtain the two-dimensional projection coordinates of the particle sample. When a certain point is larger than a preset range centered on that point, the point is determined to be the two-dimensional position of the particle sample.

6. The rapid three-dimensional localization method for particles based on local reconstruction according to claim 3, characterized in that, The acquisition of the local interferogram and z The calculation of the light intensity variation curve includes: The cutting template size is set based on the interference image of the original particle sample; Centered on the two-dimensional projection position of each particle sample, a corresponding number of local interferograms are obtained independently according to the cutting template; Numerical reconstruction is performed on each local interferogram to extract its center position, which is the two-dimensional position of the particle sample. z The light intensity change curve during directional propagation, the reconstruction distance difference between adjacent points of the light intensity change curve is consistent with the side length corresponding to the pixel unit, and the reconstruction distance range is [D1, D2] to ensure that the light intensity change curve includes the change process of particle defocus-focus-defocus.

7. A rapid three-dimensional localization method for particles based on local reconstruction according to any one of claims 1-6, characterized in that, The method for obtaining the three-dimensional trajectory is as follows: Set the maximum allowable distance for trajectory connection. If the distance between the positioning coordinates of the target in adjacent frames is less than the maximum distance value, the two points are considered to be the same sample. Arrange all 3D positioning results in chronological order. Using the positioning results of the first frame as the center, search for 3D positioning results in the second frame that satisfy the condition that the distance is less than a preset number. After judging all results, use all positioning points in the second frame as the starting point to search for corresponding points in the next frame. Repeat this process until all continuously recorded interferometric images are judged. Based on the relationship between the positioning points, the trajectory is connected frame by frame, and the positioning points that were regarded as the same sample in the previous step are arranged into a three-dimensional motion trajectory according to the time sequence.

8. A rapid three-dimensional localization system for particles based on local reconstruction, characterized in that, The system for implementing the rapid three-dimensional localization method for particles based on local reconstruction as described in any one of claims 1-7 includes the following modules: The recording module is used to record interference images of multiple particle samples in the field of view using interferometric optical imaging methods; The light field plane reconstruction module is used to evaluate the height range of the particles and perform light field reconstruction within the height range to obtain several layers of two-dimensional reconstruction plane images to reflect the spatial light field distribution within the height range. The reconstruction distance difference between adjacent planes is equal and they are uniformly distributed within the observation range. The two-dimensional positioning module is used to superimpose the light field intensity values ​​of the reconstructed planes of each layer, and obtain the two-dimensional projection coordinates of the particles by searching for local maxima in the superposition results. The local 3D light field reconstruction module is used to crop a local interferogram containing the main information of the particle sample from the original interferogram, centered on the 2D projection coordinates, and reconstruct the 2D projection coordinate position based on the local interferogram. z Curve of light intensity variation in direction; 3D positioning module, used for... z Feature extraction is performed on the light intensity variation curve along the direction to obtain the two-dimensional projection coordinates of the location. z The three-dimensional coordinates of the particles are determined by combining the coordinates with the two-dimensional positioning results. The trajectory tracking module is used to obtain the three-dimensional trajectories of multiple particles in the field of view based on the spatiotemporal continuity and the distribution of three-dimensional coordinate points.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the rapid three-dimensional localization method for microparticles based on local reconstruction as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they implement the steps of the rapid three-dimensional localization method for microparticles based on local reconstruction as described in any one of claims 1-7.

Images