Non-vision field imaging method based on area array single photon detection data co-coking

Abstract

The invention relates to the technical field of single-photon detection, in particular to a non-vision field imaging method based on area array single-photon detection data co-coking, which comprises the following steps: discretizing a detection space into coarse split elements in a space coordinate system, screening the coarse split elements based on a histogram of imaging points on a detection wall surface, and calculating the area array single-photon detection data co-coking; acquiring a hidden object positioning area represented by the rough split elements; discretizing a hidden object positioning area into subdivided voxels, and screening the subdivided voxels based on a histogram; acquiring a compensation time position and a photon count value compensation value based on the spatial position of the screened subdivided voxel; based on a compensation time position and a photon count value compensation value, performing co-coking compensation on the histogram; according to the invention, detection data processing and non-vision field imaging image high-quality recovery can be realized under actual non-ideal irradiation and detection conditions.

Description

Technical Field

[0001] The invention relates to the technical field of single-photon detection, and in particular to a non-field of view imaging method based on confocalization of area array single-photon detection data. Background Art

[0002] Non-line-of-sight imaging technology is an emerging imaging method that overcomes the limitation of traditional imaging technology that can only image objects within the line of sight, and can detect and image objects outside the line of sight. This technology mainly uses laser transient imaging technology, through which the laser emitted by the pulsed laser scatters with the object, and then the single-photon detector captures the scattered photons and records the flight time of the photons. A histogram of photon counts and flight time can be constructed, which can infer the position and shape of the object, thereby reconstructing the three-dimensional image of the object.

[0003] Non-line-of-sight imaging systems can be divided into two detection models: confocal and non-confocal. In the confocal model, the illumination point and the imaging point are at the same position, while in the non-confocal model, the two points are not at the same position. In the non-confocal model, by changing the position of the imaging point, multiple ellipsoidal surfaces can be obtained, and their intersection is the position of the object. In the confocal model, the position of the object is determined by the spherical surface. Image restoration algorithm is the key to non-line-of-sight imaging technology, among which the back-projection algorithm and the light cone transformation algorithm are the two main algorithms. The back-projection algorithm approximates the position of the object by calculating the probability distribution of the object in the object space voxels, while the light cone transformation algorithm can obtain the exact solution of the hidden object.

[0004] At present, single-point and array single-photon detectors are the two main types of detectors. Single-point detectors have complete imaging theory and high image restoration quality, but the acquisition time is long. Although the imaging quality of array detectors is relatively low, they can collect data from multiple imaging points at the same time, significantly shortening the data acquisition time and having greater practical potential. Summary of the invention

[0005] The present invention provides a non-field of view imaging method based on confocalization of area array single-photon detection data, which solves the problem in the prior art that the area array detector uses a non-confocal detection model and has low imaging quality.

[0006] The present invention provides a non-line-of-sight imaging method based on confocalization of area array single-photon detection data, comprising:

[0007] Step S1, establishing a spatial coordinate system for the detection space, discretizing the detection space into coarse voxels in the spatial coordinate system, screening the coarse voxels based on a histogram of imaging points on the detection wall, and obtaining a hidden object positioning area represented by the coarse voxels;

[0008] Step S2, discretizing the hidden object positioning area into subdivided voxels, screening the subdivided voxels based on the histogram, and obtaining the screened subdivided voxels;

[0009] Step S3, based on the spatial position of the filtered subdivided voxels, obtaining a compensation time position and a photon count value compensation value; based on the compensation time position and the photon count value compensation value, performing confocal compensation on the histogram;

[0010] Step S4: Use the compensated histogram to restore the image of the hidden object.

[0011] Preferably, step S1 specifically includes:

[0012] Step S1-1, determine the center point of the field of view of the non-confocal detection space as the origin, determine the direction of the line from the origin to the illumination point as the positive direction of the y-axis, determine the direction on the field of view plane and perpendicular to the y-axis as the x-axis direction, and determine the direction perpendicular to the field of view plane as the z-axis direction, and establish a spatial coordinate system;

[0013] Step S1-2, traversing the histogram of all imaging points on the detected wall, screening the coarse voxels, and removing invalid coarse voxels;

[0014] Step S1-3: determine the set of regions where the filtered coarse-divided voxels are located as the hidden object positioning region.

[0015] Preferably, step S1-2 specifically includes:

[0016] Step S1-21, discretizing the detection space into 32×32×32 coarse voxels;

[0017] Step S1-22, obtaining a histogram of all imaging points on the detected wall, wherein the abscissa in the histogram is the photon flight time and the ordinate is the number of photons, and obtaining the maximum and minimum values ​​of the photon flight time of the histogram;

[0018] Step S1-23, for the histogram of each imaging point, each coarse voxel is judged, including: obtaining the sum of the distances between the coarse voxel and the imaging point and the illumination point, dividing it by the speed of light to obtain the photon flight time of the coarse voxel, and if the photon flight time of the coarse voxel is not within the maximum and minimum value range of the photon flight time of the histogram, the coarse voxel is removed.

[0019] Preferably, step S2 specifically includes:

[0020] Step S2-1, discretizing each coarse voxel in the hidden object positioning area into 2×2×2 subdivided voxels;

[0021] Step S2-2: for each time position in the histogram of each imaging point, an ellipsoid with the illumination point and the imaging point as the focus and the photon flight distance corresponding to the time position as the focal length is obtained; and the subdivided voxels at the position where the ellipsoid surface falls into the hidden object positioning area are determined as the screened subdivided voxels.

[0022] Preferably, the step of determining the subdivided voxels at the position where the ellipsoid surface falls into the hidden object locating area as the screened subdivided voxels specifically includes: for each subdivided voxel, if at least one of the eight vertices of the subdivided voxel is located inside the ellipsoid and at least one of the 8 vertices is located outside the ellipsoid, then it is determined that the subdivided voxel does not need to be eliminated; otherwise, it is determined that the subdivided voxel needs to be eliminated; and a set of all subdivided voxels that do not need to be eliminated is used as the screened subdivided voxels.

[0023] Preferably, step S3 specifically includes:

[0024] Step S3-1, obtaining the compensation time position in the histogram based on the distance between the filtered subdivided voxels and the confocal point; the confocal point is the midpoint between the imaging point and the illumination point;

[0025] Step S3-2, obtaining a compensation value of the photon count value according to the distance between the screened subdivided voxel and the illumination point, the distance between the screened subdivided voxel and the imaging point, and the distance between the screened subdivided voxel and the confocal point;

[0026] Step S3-3, compensating the histogram based on the compensation time position and the compensation value of the photon count value until the compensation of the histograms of all imaging points is completed, and the histograms of all imaging points after the final compensation are used as the confocal detection data.

[0027] Preferably, step S3-1 specifically includes:

[0028] The distance from the subdivided voxel point to the confocal point is calculated, and the time position to be compensated in the histogram is obtained from the distance from the subdivided voxel point to the confocal point; the distance from the subdivided voxel point to the confocal point is calculated by the following formula:

[0029]

[0030] Wherein, n represents the distance between the subdivided voxel point and the imaging point, m represents the distance between the imaging point and the illumination point, p represents the distance between the subdivided voxel point and the illumination point, and q represents the distance between the subdivided voxel point and the confocalization point.

[0031] Preferably, step S3-2 specifically includes:

[0032] Obtaining a first attenuation of the photon count value from the distance between the subdivided voxel and the illumination point and the distance between the subdivided voxel and the imaging point;

[0033] A second attenuation of the photon count value is obtained according to the distance between the subdivided voxel and the confocal point;

[0034] A compensation value of the photon count value is obtained according to the first attenuation and the second attenuation of the photon count value.

[0035] Preferably, the step S3-3 of compensating the histogram based on the compensation time position and the compensation value of the photon count value specifically includes:

[0036] Based on the compensation value of the photon count value, the photon count value at the compensation time position in the histogram is compensated.

[0037] Compared with the prior art, the present invention has at least the following beneficial effects:

[0038] The present invention provides a method for confocalizing the collected data and realizing confocal detection data simulation for area array single photon detection, including a rapid positioning process and a data correction confocalization process, wherein the rapid positioning process accurately limits the effective distribution position of the object to the accurate area of ​​the subdivision from coarse to fine by coarse and fine division of voxels, and the local ellipsoid interpolation process provides a method for converting non-confocal detection data into confocal detection data, and the confocalization accuracy can be improved through these two processes. The method provided by the present invention can be flexibly applied to non-field of view imaging in a variety of practical scenarios, and can realize detection data processing and high-quality image restoration under actual non-ideal illumination and detection conditions, and promote the practical application of non-field of view imaging technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The drawings are only for the purpose of illustrating particular embodiments and are not to be construed as limiting the invention.

[0040] Figure 1 The present invention discloses a flow chart of a non-field of view imaging method based on confocalization of area array single-photon detection data.

[0041] Figure 2a It is a schematic diagram of the viewing area and non-viewing area range disclosed in the present invention.

[0042] Figure 2b It is a schematic diagram of the non-line-of-sight imaging system disclosed in the present invention.

[0043] Figure 3a This is a schematic diagram of a simplified model of non-confocal detection disclosed in the present invention.

[0044] Figure 3b This is a schematic diagram of a simplified model of confocal detection disclosed in the present invention.

[0045] Figure 4 The figure is a flow chart of the confocalization algorithm disclosed in the present invention.

[0046] Figure 5 This is a schematic diagram of the positioning of the rapid positioning module disclosed in the present invention.

[0047] Figure 6 Schematic diagram of voxel subdivision of the positioning area disclosed in the present invention

[0048] Figure 7 This is a schematic diagram of the principle of compensating photon count values ​​disclosed in the present invention.

[0049] Figure 8 It is a top view of the ellipsoid surface disclosed in the present invention falling into the hidden object positioning area.

[0050] Fig. 9 A schematic diagram of a target object for a non-visual field experiment disclosed in the present invention.

[0051] Fig.10a This is a schematic diagram of the reconstruction result of the letter “E” disclosed in the present invention.

[0052] Fig.10b A row schematic diagram is selected for the reconstruction result of the letter "E" disclosed in the present invention.

[0053] Fig.11a This is the albedo distribution diagram of the confocalization algorithm disclosed in the present invention.

[0054] Fig.11b This is the albedo distribution diagram of the back-projection algorithm disclosed in the present invention. DETAILED DESCRIPTION

[0055] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. In addition, the present invention can also be implemented in other ways different from those described herein, and therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0056] The present invention is applied to non-line-of-sight imaging systems, such as Figure 2a As shown in the figure, the area outside the field of view is generally called the non-field of view range. The non-field of view imaging technology can reconstruct the three-dimensional image of the non-field of view target, so that the detector can observe the target at a certain distance from the obstacle. The schematic diagram is shown in Figure 2b The working process of the non-line-of-sight imaging system is as follows: the laser emitted by the pulsed laser scatters with the object, and then the single-photon detector captures the scattered photons and records the photon flight time. A histogram of photon counts and flight time can be constructed to infer the position and shape of the object, thereby reconstructing the three-dimensional image of the object.

[0057] Generally, according to the relative position of the illumination point and the detection point on the intermediate wall, the detection model is divided into two types: confocal and non-confocal. In the confocal model, the illumination point and the imaging point are at the same position, while in the non-confocal model, the two points are not at the same position. Figure 3a , Figure 3b The simplified model diagrams of non-confocal detection and confocal detection are shown respectively. In the non-confocal model, by changing the position of the imaging point, multiple ellipsoidal surfaces can be obtained, and their intersection is the position of the object. In the confocal model, the object position is determined by the focus of the spherical surface.

[0058] In order to combine the high efficiency of the area array detector and the high imaging quality of the confocal model, the method provided by the present invention can convert the non-confocal data collected by the area array detector into confocal data, thereby utilizing the imaging algorithm of the confocal detection model. This method helps to improve the imaging quality of non-confocal detection data, making it more suitable for practical application scenarios and promoting the practical application of non-field of view imaging technology.

[0059] In order to illustrate the effectiveness of the method proposed by the present invention, the technical solution of the present invention is described in detail below through a specific embodiment. Figure 1 The present invention provides a non-line-of-sight imaging method based on confocalization of area array single-photon detection data, comprising the following steps:

[0060] Step S1, establishing a spatial coordinate system for the detection space, discretizing the detection space into coarse voxels in the spatial coordinate system, screening the coarse voxels based on a histogram of imaging points on the detection wall, and obtaining a hidden object positioning area represented by the coarse voxels;

[0061] The confocalization algorithm based on local ellipsoid interpolation of the present invention mainly includes two modules, namely a fast positioning module and an ellipsoid interpolation module. The algorithm flow of the present invention is as follows: Figure 4 As shown in the figure, the first module mainly realizes the positioning of hidden objects, and obtains a three-dimensional cube with known coordinates, and the hidden object is in the cube.

[0062] The idea of ​​implementing the fast positioning method is as follows:

[0063] (1) Establish a spatial rectangular coordinate system, where the origin of the coordinate system is the center of the detection field of view in the non-confocal case; the positive direction of the y-axis is the direction of the line from the origin to the illumination point, which is the direction of column increase; the x-axis is perpendicular to the y-axis on the field of view plane, and the positive direction is the direction of row increase; the z-axis is perpendicular to the field of view and conforms to the right-hand rule with the x- and y-axes. Discretize the object space into grid_x, grid_y, and grid_z. The degree of discretization does not need to be too fine. It is only necessary to determine the approximate spatial range of the hidden object. For example, discretize the object space into a coarse voxel distribution of 32×32×32.

[0064] (2) Take out the histogram of an imaging point on the detected wall and select the maximum value T of its useful data max With the minimum value T min , the coordinate values ​​of the imaging point are grid_fov_x, grid_fov_y, and the coordinate values ​​of the lighting point are source_x, source_y.

[0065] (3) Calculate the sum of the distances from the coarse voxels in the discrete object space to the imaging point and the illumination point, and divide it by the speed of light to convert it into the photon flight time T.

[0066] (4) Screen the photon flight time T of each coarse-scale voxel and select those that are not within T min and T max The coarse voxels in between are removed.

[0067] (5) Traversing the histograms of several imaging points, the distribution range of the remaining coarse voxel coordinate values ​​can represent the distribution range of the hidden object.

[0068] The spatial range of hidden objects located by the fast positioning module is as follows: Figure 5 The remaining coarse-divided voxel coordinate value distribution range finally obtained by the fast positioning module is input into the confocalization calculation module as the positioning area as the initial positioning information, which can improve the calculation speed of the algorithm while ensuring the confocalization performance.

[0069] In some embodiments, step S1 includes:

[0070] Step S1-1, determine the center point of the field of view of the non-confocal detection space as the origin, determine the direction of the line from the origin to the illumination point as the positive direction of the y-axis, determine the direction on the field of view plane and perpendicular to the y-axis as the x-axis direction, and determine the direction perpendicular to the field of view plane as the z-axis direction, and establish a spatial coordinate system;

[0071] Step S1-2, traversing the histogram of all imaging points on the detected wall, screening the coarse voxels, and removing invalid coarse voxels;

[0072] Step S1-3: determine the set of regions where the filtered coarse-divided voxels are located as the hidden object positioning region.

[0073] In some embodiments, step S1-2 includes:

[0074] Step S1-21, discretizing the detection space into 32×32×32 coarse voxels;

[0075] Step S1-22, obtaining a histogram of all imaging points on the detected wall, wherein the abscissa in the histogram is the photon flight time and the ordinate is the number of photons, and obtaining the maximum and minimum values ​​of the photon flight time of the histogram;

[0076] Step S1-23, for the histogram of each imaging point, each coarse voxel is judged, including: obtaining the sum of the distances between the coarse voxel and the imaging point and the illumination point, dividing it by the speed of light to obtain the photon flight time of the coarse voxel, and if the photon flight time of the coarse voxel is not within the maximum and minimum value range of the photon flight time of the histogram, the coarse voxel is removed.

[0077] The local ellipsoid interpolation-based confocalization algorithm of the present invention mainly includes two modules, namely a fast positioning module and an ellipsoid interpolation module. Among them, the second module mainly realizes the conversion of non-confocal detection data into confocal detection data through the newly defined confocalization point and the positioning result of the fast positioning module.

[0078] The essential difference between the non-confocal detection model and the confocal detection model lies in the relative position of the illumination point and the imaging point. Figure 7 As shown in the figure, the confocalization theoretical model is to select a new illumination point (imaging point) outside the illumination point and the imaging point under the non-confocal detection model, and then select an optimal spherical radius so that the sphere with the new illumination point (imaging point) as the center can achieve the best fit with the original ellipsoid model.

[0079] In the traditional confocalization process, it is assumed that the distribution probability of hidden objects on the ellipsoid is equal. However, in the actual non-viewing scene, hidden objects are only distributed in a part of the ellipsoid surface, so how to accurately limit the effective distribution position of the object to a certain accurate area is more important for improving the confocalization accuracy. In view of the above problems existing in confocalization, the present invention provides a confocalization algorithm based on local ellipsoid interpolation to optimize the problem and improve the confocalization accuracy.

[0080] Step S2: discretize the hidden object positioning area into subdivided voxels, filter the subdivided voxels based on the histogram, and obtain the filtered subdivided voxels.

[0081] Step S3, based on the spatial position of the filtered subdivided voxels, obtaining a compensation time position and a photon count value compensation value; based on the compensation time position and the photon count value compensation value, performing confocal compensation on the histogram.

[0082] The implementation idea of ​​the local ellipsoid interpolation method provided by the present invention is as follows:

[0083] (1) Define the origin of the coordinate system as the center of the detection field of view in the non-confocal case; the positive direction of the y-axis is the direction of the line from the origin to the irradiation point, which is the direction of increasing columns; the x-axis is perpendicular to the y-axis on the plane of the field of view, and its positive direction is the direction of increasing rows; the z-axis is perpendicular to the field of view and conforms to the right-hand rule with the x-axis and the y-axis. The coordinate system in this step is the same as the coordinate system in step S1.

[0084] (2) Voxel refinement is performed on the object space region obtained by the positioning module, such as Figure 6 As shown in Figure 1, each voxel point in the localization area is discretized into 2×2×2 subdivided voxels, and this distribution is used for the confocalization module. Select a detection point and obtain the original histogram data of the detection point, where T timebin is the original timing accuracy, T new Timing precision for confocalization.

[0085] (3) Obtain a time position in the original histogram data, and take the object space located in the coordinate system as a reference to obtain an ellipsoid with the illumination point and the imaging point as the focus and the photon flight distance corresponding to the time position as the focal length; determine the subdivided voxels at the position where the ellipsoid surface falls into the hidden object positioning area as the screened subdivided voxels; then, obtain the distance from each subdivided voxel in the hidden object positioning area to the confocal point.

[0086] (4) Select a subdivided voxel position, calculate the distance from the subdivided voxel point to the confocal point, and obtain the time position to be compensated in the histogram based on the distance from the subdivided voxel point to the confocal point; calculate the distance from the subdivided voxel point to the confocal point by the following formula:

[0087]

[0088] Wherein, n represents the distance between the subdivided voxel point and the imaging point, m represents the distance between the imaging point and the illumination point, p represents the distance between the subdivided voxel point and the illumination point, and q represents the distance between the subdivided voxel point and the confocalization point.

[0089] (5) Compensate the photon count value by the midline theorem. Specifically, it includes: obtaining the first attenuation of the photon count value by the distance between the subdivided voxel and the illumination point and the distance between the subdivided voxel and the imaging point; obtaining the second attenuation of the photon count value by the distance between the subdivided voxel and the confocal point; obtaining the compensation value of the photon count value by the first attenuation and the second attenuation of the photon count value. Based on the compensation value of the photon count value, compensate the photon count value at the compensation time position in the histogram. Go back to step (4), and follow the process loop from (4) to (5) to complete the calculation of all subdivided voxel points, such as Figure 8 shown.

[0090] (6) Return to step (3) and repeat the process from (3) to (5) to complete the calculation of all subdivided voxel points corresponding to all time positions.

[0091] (7) Return to step (2) and repeat the process from (2) to (6) to complete the calculation of all detection points on the intermediate wall.

[0092] In some embodiments, Figure 8As shown, the step of determining the subdivided voxels at the position where the ellipsoid surface falls into the hidden object positioning area as the screened subdivided voxels specifically includes: for each subdivided voxel, if at least one of the eight vertices of the subdivided voxel is located inside the ellipsoid and at least one vertex is located outside the ellipsoid, then it is determined that the subdivided voxel does not need to be eliminated; otherwise, it is determined that the subdivided voxel needs to be eliminated; and a set of all subdivided voxels that do not need to be eliminated is used as the screened subdivided voxels.

[0093] In some embodiments, the present invention provides a non-line-of-sight imaging method based on confocalization of area array single-photon detection data, further comprising:

[0094] Step S4: Use the confocal detection data and an image restoration algorithm based on light cone transformation to restore the image of the hidden object.

[0095] In some embodiments, the confocalization algorithm based on local ellipsoid interpolation outputs a confocalization histogram, which is input as input data into the light cone transformation image reconstruction algorithm. In this way, a closed solution of the hidden object can be solved with a faster reconstruction time, solving the problem that the probability distribution of the hidden object can only be estimated using the back projection algorithm for non-confocal detection data in the past.

[0096] In real experiments, common "E"-shaped objects are used to conduct 3D reconstruction image performance verification experiments, such as Fig. 9 As shown, the three lateral widths of the object are set to 8 cm, and the interval between two adjacent lateral widths is 10 cm. The hidden object is placed in the center of the virtual field of view after confocalization to evaluate the effect of confocal reconstruction of the image.

[0097] Under the dark light condition of the laboratory, the source data was obtained using the area array single photon detection non-field of view imaging system. First, the image of the hidden object was reconstructed using the back-projection algorithm most commonly used in non-confocal detection configuration as a control group. Then, the non-confocal detection data was confocalized using the traditional NMO correction algorithm and the confocalization algorithm based on local ellipsoid interpolation proposed in this invention, and the image was reconstructed using the light cone transformation reconstruction algorithm. The reconstruction results are shown in Figure 2. Fig.10a shown.

[0098] In order to evaluate the image restoration level of each algorithm for the "E"-shaped object, a row of pixels in the plane albedo distribution of the hidden object obtained by image restoration was selected for evaluation. The row passes through the three vertical bars of the letter E and the interval between the vertical bars. The number of selected rows and the true albedo of the small "E" under the row and the albedo distribution under different algorithms are shown in the figure below. Fig.10b , Fig.11a , Fig.11b shown.

[0099] The albedo difference between the first vertical bar and the first interval of the small "E" in the image restored by the confocalization algorithm reaches 0.16, and the back-projected albedo distribution map cannot distinguish the vertical bar from the interval. Among the above two algorithms, only the confocalization algorithm achieves the albedo at the interval being less than the albedo of the two adjacent vertical bars, achieving the purpose of resolving the hidden object, while the back-projection algorithm cannot achieve the purpose of resolving the hidden object. Based on the experimental results of confocalization of non-field of view imaging of comprehensive array single-photon detection, the confocalization method based on local ellipsoid interpolation verifies the feasibility and superiority of the confocalization method proposed in the present invention.

[0100] Although the specific embodiments of the present invention have been described in a specific order, each action or step should be understood to require that such action or step be performed in the specific order shown or in a sequential order, or require that all illustrated actions or steps should be performed to obtain the desired result. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Certain features described in the context of a separate embodiment can also be implemented in a single implementation in combination. On the contrary, the various features described in the context of a single implementation can also be implemented in multiple implementations individually or in any suitable sub-combination.

[0101] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A non-line-of-sight imaging method based on confocalization of area array single-photon detection data, comprising the following steps: Step S1, establishing a spatial coordinate system for the detection space, discretizing the detection space into coarse voxels in the spatial coordinate system, screening the coarse voxels based on a histogram of imaging points on the detection wall, and obtaining a hidden object positioning area represented by the coarse voxels; Step S2, discretizing the hidden object positioning area into subdivided voxels, screening the subdivided voxels based on the histogram, and obtaining the screened subdivided voxels; Step S3, based on the spatial position of the filtered subdivided voxels, obtaining a compensation time position and a photon count value compensation value; based on the compensation time position and the photon count value compensation value, performing confocal compensation on the histogram; Step S4: Use the compensated histogram to restore the image of the hidden object.

2. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 1, characterized in that: Step S1 specifically includes: Step S1-1, determine the center point of the field of view of the non-confocal detection space as the origin, determine the direction of the line from the origin to the illumination point as the positive direction of the y-axis, determine the direction on the field of view plane and perpendicular to the y-axis as the x-axis direction, and determine the direction perpendicular to the field of view plane as the z-axis direction, and establish a spatial coordinate system; Step S1-2, discretizing the detection space, traversing the histogram of all imaging points on the detection wall, screening the coarse voxels, and removing invalid coarse voxels; Step S1-3: determine the set of regions where the filtered coarse-divided voxels are located as the hidden object positioning region.

3. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 2, characterized in that: Step S1-2 specifically includes: Step S1-21, discretizing the detection space into 32×32×32 coarse voxels; Step S1-22, obtaining a histogram of all imaging points on the detected wall, wherein the abscissa in the histogram is the photon flight time and the ordinate is the number of photons, and obtaining the maximum and minimum values ​​of the photon flight time of the histogram; Step S1-23, for the histogram of each imaging point, each coarse voxel is judged, including: obtaining the sum of the distances between the coarse voxel and the imaging point and the illumination point, dividing it by the speed of light to obtain the photon flight time of the coarse voxel, and if the photon flight time of the coarse voxel is not within the maximum and minimum value range of the photon flight time of the histogram, the coarse voxel is removed.

4. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 3, characterized in that: Step S2 specifically includes: Step S2-1, discretizing each coarse voxel in the hidden object positioning area into 2×2×2 subdivided voxels; Step S2-2: for each time position in the histogram of each imaging point, an ellipsoid with the illumination point and the imaging point as the focus and the photon flight distance corresponding to the time position as the focal length is obtained; and the subdivided voxels at the position where the ellipsoid surface falls into the hidden object positioning area are determined as the screened subdivided voxels.

5. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 4, characterized in that: The step of determining the subdivided voxels at the positions where the ellipsoid surface falls into the hidden object positioning area as the screened subdivided voxels specifically includes: For each subdivided voxel, if at least one of the eight vertices of the subdivided voxel is located inside the ellipsoid and at least one of the vertices is located outside the ellipsoid, it is determined that the subdivided voxel does not need to be eliminated; otherwise, it is determined that the subdivided voxel needs to be eliminated; The set of all subdivided voxels that do not need to be eliminated is taken as the screened subdivided voxels.

6. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 5, characterized in that: Step S3 specifically includes: Step S3-1, obtaining the compensation time position in the histogram based on the distance between the filtered subdivided voxels and the confocal point; the confocal point is the midpoint between the imaging point and the illumination point; Step S3-2, obtaining a compensation value of the photon count value according to the distance between the screened subdivided voxel and the illumination point, the distance between the screened subdivided voxel and the imaging point, and the distance between the screened subdivided voxel and the confocal point; Step S3-3, compensating the histogram based on the compensation time position and the compensation value of the photon count value until the compensation of the histograms of all imaging points is completed, and the histograms of all imaging points after the final compensation are used as the confocal detection data.

7. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 6, characterized in that: Step S3-1 specifically includes: The distance from the subdivided voxel point to the confocal point is calculated, and the time position to be compensated in the histogram is obtained from the distance from the subdivided voxel point to the confocal point; the distance from the subdivided voxel point to the confocal point is calculated by the following formula: Wherein, n represents the distance between the subdivided voxel point and the imaging point, m represents the distance between the imaging point and the illumination point, p represents the distance between the subdivided voxel point and the illumination point, and q represents the distance between the subdivided voxel point and the confocalization point.

8. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 7, characterized in that: Step S3-2 specifically includes: Obtaining a first attenuation of the photon count value from the distance between the subdivided voxel and the illumination point and the distance between the subdivided voxel and the imaging point; A second attenuation of the photon count value is obtained according to the distance between the subdivided voxel and the confocal point; A compensation value of the photon count value is obtained according to the first attenuation and the second attenuation of the photon count value.

9. The non-line-of-sight imaging method based on confocalization of area array single-photon detection data according to claim 8, characterized in that: The step S3-3 of compensating the histogram based on the compensation time position and the compensation value of the photon count value specifically includes: Based on the compensation value of the photon count value, the photon count value at the compensation time position in the histogram is compensated.

Images