Single photon reconstruction method and device and electronic equipment

Abstract

The invention provides a single photon reconstruction method and device and electronic equipment, and relates to the technical field of single photon imaging, and the method comprises the steps: obtaining a photon distribution observation value, and constructing a photon counting model according to the photon distribution observation value; a negative Poisson maximum likelihood function is used as a loss function, the photon counting model is optimized, and an optimized constraint form is determined; determining a sub-problem of a photon counting model according to the loss function and the optimized constraint form; carrying out secondary iteration solution on the sub-problems to obtain photon distribution under the target ideal condition; according to the method, the depth image and the intensity image are generated according to the photon distribution under the target ideal condition, so that short-distance and long-distance adaptive rapid imaging can be realized, the denoising effect can be effectively improved, and more accurate three-dimensional space distribution of photons can be statistically realized.

Description

Technical Field

[0001] The present application relates to the field of single-photon imaging technology, and in particular to a single-photon reconstruction method, device and electronic equipment. Background Art

[0002] Typically, single-photon data has the characteristics of a large number of noise photons and sparse signal photons. Existing single-photon imaging methods mainly include the traditional sparse Poisson reconstruction method, which can only be used for close-range two-dimensional imaging and cannot solve the distribution of photons in the time domain and the point spread function matrix of photons. Summary of the invention

[0003] In order to at least overcome the above-mentioned deficiencies in the prior art, the purpose of the present application is to provide a single-photon reconstruction method, device and electronic device.

[0004] In a first aspect, an embodiment of the present application provides a single-photon reconstruction method, the single-photon reconstruction method comprising:

[0005] Acquire photon distribution observation values, and construct a photon counting model according to the photon distribution observation values;

[0006] Using a negative Poisson maximum likelihood function as a loss function, optimizing the photon counting model, and determining an optimized constraint form;

[0007] Determining a subproblem of the photon counting model according to the loss function and the optimized constraint form;

[0008] Solve the sub-problem twice iteratively to obtain the photon distribution under the target ideal;

[0009] A depth image and an intensity image are generated according to the ideal photon distribution of the target.

[0010] In a possible implementation, the step of using a negative Poisson maximum likelihood function as a loss function to optimize the photon counting model and determining an optimized constraint form includes:

[0011] According to the photon counting model, a probability function of photons is obtained;

[0012] A negative Poisson maximum likelihood function is used as the loss function, and an optimized constraint form is obtained according to the probability function of the photon.

[0013] In a possible implementation, the probability function of the photon is:

[0014]

[0015] Among them, e irepresents the standard unit basis vector, y represents the observed value of the photon distribution, A represents the point spread function of the photon, f represents the photon distribution under the target ideal, and T represents the transpose;

[0016] The loss function is:

[0017]

[0018] Among them, ε is a constant not equal to 0;

[0019] The optimized constraint form is:

[0020] minimizeΦ(f)≡F(f)+τpen(f)

[0021] subjecttof≥0

[0022] Among them, τ represents a constant and pen represents a penalty term.

[0023] In a possible implementation, the step of determining the subproblem of the photon counting model according to the loss function and the optimized constraint form includes:

[0024] constructing an approximate expression according to the optimized constraint form, and using the difference as a penalty term of the approximate expression;

[0025] Use the Taylor function to perform a quadratic expansion on the loss function to obtain a Taylor expansion;

[0026] The subproblems of the photon counting model are obtained according to the approximate expression and the Taylor expansion.

[0027] In a possible implementation, the approximate expression constructed according to the optimized constraint form is:

[0028]

[0029] subjecttof≥0

[0030] Where f represents the photon distribution under the ideal target, k represents the number of iterations, represents the approximate value of the loss function; τ represents a constant; pen represents a penalty term;

[0031] The Taylor expansion is:

[0032]

[0033] Among them, α k I represents the approximate value of the second-order derivative of the loss function, I represents the identity matrix, and F(f k ) represents the loss function, represents the first-order derivative of the loss function;

[0034] The subproblems of the photon counting model are:

[0035]

[0036] subjecttof≥0,

[0037] in,

[0038] In a possible implementation, the photon counting model is:

[0039] y=Af

[0040] Among them, y represents the observed value of photon distribution, A represents the point spread function of photons, and f represents the photon distribution under the ideal target.

[0041] In a possible implementation manner, after the step of constructing a photon counting model according to the photon distribution observation value, the method further includes:

[0042] Acquiring experimental equipment parameters, and calculating the point spread function of photons according to the experimental equipment parameters;

[0043] The point spread function is simplified in the form of a Gaussian function to obtain the probability distribution of photons in space.

[0044] In a possible implementation, the point spread function is:

[0045]

[0046] Among them, σ x , σ y , σ z represents the standard deviation, (x, y, z) represents the spatial coordinate position of the photon, and I0 represents a constant;

[0047] The probability distribution of photons in space is:

[0048]

[0049] Among them, (a, b, c) represents the three-dimensional spatial position of the photon, t represents the time range of the photon,

[0050]

[0051]

[0052] Spot represents the spot radius, fov represents the radius of the receiving field of view, and d represents the imaging distance.

[0053] In a second aspect, the present application also provides a single-photon reconstruction device, including:

[0054] A construction module is used to obtain photon distribution observation values ​​and construct a photon counting model according to the photon distribution observation values;

[0055] An optimization module, used to optimize the photon counting model using a negative Poisson maximum likelihood function as a loss function, and determine an optimized constraint form;

[0056] A determination module, used to determine a sub-problem of the photon counting model according to the loss function and the optimized constraint form;

[0057] A secondary iterative solution module is used to perform secondary iterative solutions to the sub-problems to obtain the photon distribution under the target ideal condition;

[0058] A generation module is used to generate a depth image and an intensity image according to the ideal photon distribution of the target.

[0059] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by the processor, the method described in any of the above aspects is implemented.

[0060] Based on any of the above aspects, the single-photon reconstruction method, device and electronic device provided in the embodiments of the present application can obtain the photon distribution under the ideal target by constructing a photon counting model and iteratively solving the photon counting model. Compared with the existing technology, it can achieve adaptive rapid imaging at close and long distances, and can effectively improve the denoising effect, and statistically achieve a more accurate three-dimensional spatial distribution of photons. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0062] Figure 1 One of the flow charts of the single photon reconstruction method provided in the embodiment of the present application;

[0063] Figure 2 A schematic diagram of sub-steps of step S120 provided in an embodiment of the present application;

[0064] Figure 3 A schematic diagram of sub-steps of step S130 provided in an embodiment of the present application;

[0065] Figure 4 A schematic diagram of a second iteration provided in an embodiment of the present application;

[0066] Figure 5 The second flowchart of the single photon reconstruction method provided in the embodiment of the present application;

[0067] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0068] Figure 7 A schematic diagram of the structure of a single-photon reconstruction device provided in an embodiment of the present application.

[0069] Icons: 800 - electronic device; 810 - processor; 820 - machine-readable storage medium; 830 - single-photon reconstruction device; 831 - construction module; 832 - optimization module; 833 - determination module; 834 - secondary iterative solution module; 835 - generation module. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0071] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0072] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0073] In the description of this application, it should be noted that the terms "upper", "lower", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product of the application is usually placed when in use, which is only for the convenience of describing this application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0074] It should be noted that, in the absence of conflict, different features in the embodiments of the present application may be combined with each other.

[0075] The specific implementation methods of the present application are described in detail below in conjunction with the accompanying drawings.

[0076] Please refer to Figure 1 , Figure 1 A schematic flow chart of a single-photon reconstruction method provided by the present embodiment is given as an example. The method may include the following steps.

[0077] Step S110, obtaining photon distribution observation values, and constructing a photon counting model according to the photon distribution observation values.

[0078] In this embodiment, the photon distribution observation value can be obtained through experimental equipment, and the photon counting model can be constructed using a non-uniform Poisson distribution model based on the photon distribution observation value. The photon distribution observation value is the photon distribution observed by the experimental equipment, and can include the distribution of signal photons, white noise photons, dark current photons, etc.

[0079] The photon counting model constructed may be:

[0080] y=Af

[0081] Among them, y represents the observed value of the photon distribution, A represents the point spread function of the photon, which represents the probability distribution of the photon in three dimensions of space, and f represents the photon distribution under the ideal target.

[0082] Both y and A are known numbers, and y, A and f are all non-negative numbers. Therefore, the sparse characteristics can be used to perform optimization iterations to obtain the ideal photon distribution.

[0083] Step S120, using the negative Poisson maximum likelihood function as a loss function, optimizing the photon counting model, and determining an optimized constraint form.

[0084] In this embodiment, the photon counting model constructed in step S110 may be optimized based on the idea of ​​maximum likelihood function to obtain the optimized constraint form.

[0085] Step S130, determining the sub-problems of the photon counting model according to the loss function and the optimized constraint form.

[0086] In this embodiment, the sub-problem of the photon counting model can be obtained according to the loss function obtained in step S120 and the optimized constraint form.

[0087] Step S140, performing a second iteration to solve the sub-problem to obtain the photon distribution under the target ideal.

[0088] In this embodiment, the sub-problem obtained in step S130 may be solved by a second iteration to obtain the photon distribution under the target ideal.

[0089] Step S150, generating a depth image and an intensity image according to the ideal photon distribution of the target.

[0090] In this embodiment, the depth image and the intensity image may be generated respectively according to the ideal photon distribution of the target obtained in step S140.

[0091] In the above design, by constructing the photon counting model and iteratively solving the photon counting model, the photon distribution under the ideal target can be obtained. Compared with the existing technology, adaptive fast imaging at close and long distances can be achieved, and the denoising effect can be effectively improved, and a more accurate three-dimensional spatial distribution of photons can be achieved statistically.

[0092] In addition, the single-photon reconstruction method provided in this embodiment can be applied to multiple fields such as medical detection, weather environment detection, and military.

[0093] In one possible implementation, see Figure 2 , step S120 may include the following sub-steps.

[0094] Step S121, obtaining a probability function of photons according to the photon counting model.

[0095] In this embodiment, since the photon distribution obeys the Poisson distribution, the probability function of the photon is:

[0096]

[0097] Among them, e i represents the standard unit basis vector, y represents the observed value of the photon distribution, A represents the point spread function of the photon, f represents the photon distribution under the target ideal, and T represents the transpose.

[0098] Step S122, using the negative Poisson maximum likelihood function as the loss function, and obtaining an optimized constraint form according to the probability function of the photon.

[0099] In this embodiment, the negative Poisson maximum likelihood function can be used as the loss function, since log(y i ! ) is a constant and does not affect the optimization target, so it can be ignored. The loss function can be:

[0100]

[0101] Among them, y represents the observed value of the photon distribution, A represents the point spread function of the photon, f represents the photon distribution under the ideal target, T represents the transpose, ε is a constant not equal to 0, and the value of ε is very small, which can be used to avoid f being equal to 0.

[0102] The first derivative of the loss function and the second derivative It can be expressed as:

[0103]

[0104]

[0105] Therefore, the optimized constraint form can be:

[0106] minimizeΦ(f)≡F(f)+τpen(f)

[0107] subject to f≥0

[0108] Among them, τ represents a constant; pen represents the penalty term for non-smooth and non-convex.

[0109] In one possible implementation, see Figure 3 , step S130 may include the following sub-steps.

[0110] Step S131, constructing an approximate expression according to the optimized constraint form, and using the difference as a penalty term of the approximate expression.

[0111] In this embodiment, it is possible to use Approximately express the loss function F(f), the approximate expression constructed according to the optimized constraint form can be:

[0112]

[0113] subjecttof≥0

[0114] Where f represents the photon distribution under the ideal target, k represents the number of iterations, Represents the approximate value of the loss function; τ represents a constant; pen represents a penalty term.

[0115] In a sparse point cloud, there may be a lot of blanks in some areas, resulting in uneven information distribution, and in some areas, the point cloud may be evenly distributed. Therefore, the difference can be used as the penalty term of the approximate expression, and the expression of the difference penalty term can be:

[0116]

[0117] Among them, (x, y, z) represents the spatial position of the point cloud, z represents the distribution on the time axis, and the spatial difference on the time axis is accumulated.

[0118] Step S132, using the Taylor function to perform a second expansion on the loss function to obtain a Taylor expansion.

[0119] In this embodiment, the Hessian matrix (second-order derivative) can be approximately expressed as α k I, I represents the identity matrix, and the Taylor expansion can be:

[0120]

[0121] Among them, F(f k ) represents the loss function, Represents the first-order derivative of the loss function.

[0122] Step S133, obtaining a sub-problem of the photon counting model according to the approximate expression and the Taylor expansion.

[0123] In this embodiment, the sub-problems of the photon counting model may be:

[0124]

[0125] subjecttof≥0,

[0126] in,

[0127] Please refer to Figure 4 , during the iteration process, you can first get the initialization value f 0 , then a gradient descent algorithm (e.g., the Barzilai-Borwein method) can be used to α k Initialize and use δf k =f k -f k-1 Calculate α k :

[0128]

[0129] Here, ∈ represents a constant.

[0130] In the calculation of α k Afterwards, f can be calculated by k+1 :

[0131]

[0132] subjecttof≥0

[0133] And recalculate α k =ηα k , judge α k Is it within the set range? If not, recalculate f k+1 and α k , so that α k Repeat until α k ∈[α min ,α max ]. Where η represents a factor coefficient greater than 1, 0≤α min ≤α max .

[0134] When passing α k =ηα k The calculated α k ∈[α min ,α max ], we can judge whether the loss function is optimized to the minimum value, that is:

[0135]

[0136] Wherein, M represents a non-negative integer, σ represents a constant, σ∈(0,1).

[0137] If the loss function is optimized to the minimum value, the iteration is stopped. If the loss function is not optimized to the minimum value, the iteration is continued.

[0138] In the above design, for single-photon imaging, the secondary iterative optimization design can solve the problem quickly and accurately, and has a faster running speed than general traditional algorithms.

[0139] In one possible implementation, see Figure 5 After constructing the photon counting model according to the photon distribution observation values, the single photon reconstruction method may further include the following steps.

[0140] Step S210, obtaining experimental equipment parameters, and calculating the point spread function of photons according to the experimental equipment parameters.

[0141] In this embodiment, the experimental equipment parameters may be obtained, and the point spread function of photons may be calculated according to the experimental equipment parameters. The point spread function may be:

[0142]

[0143] Among them, σ x , σ y , σ z represents the standard deviation, which can indicate the diffusion degree of photon distribution, (x, y, z) represents the spatial coordinate position of the photon, and I0 represents a constant.

[0144] The experimental equipment parameters may include laser wavelength, laser pulse width, average power, system jitter, actual spot radius, radius of receiving field angle, scanning step length, etc.

[0145] When it is necessary to achieve clear 1.5 times diffraction-limited super-resolution imaging at a distance of 1.5km, the laser wavelength can be 1550nm, the laser pulse width can be 2.3ps, the average power can be 80mW, the system jitter can be 70ps, the actual spot radius can be 0.1876m, the radius of the receiving field of view angle can be 0.0936m, the device scanning method can adopt sub-pixel scanning imaging method, and the scanning step size can be 1 / 4FoV.

[0146] Step S220: simplify the point spread function in the form of a Gaussian function to obtain the probability distribution of photons in space.

[0147] In this embodiment, the point spread function of photons calculated at different distances and different experimental equipment parameters is different. Therefore, an adaptive convolution kernel can be designed according to parameters such as FoV, and the distribution of photons in space can be simulated in the form of a Gaussian function. The probability distribution of photons in space can be:

[0148]

[0149] Among them, the spatial range of photon distribution is [fovx, fovy, t], which represents the range in space and time respectively, (a, b, c) represents the three-dimensional spatial position of the photon, a∈[1,fovx], b∈[1,fovy], c∈[1,t].

[0150]

[0151]

[0152]

[0153] Spot represents the spot radius, fov represents the radius of the receiving field of view, and d represents the imaging distance.

[0154] In the above design, by designing an adaptive convolution kernel, it is possible to be compatible with photon distributions at various distances and achieve accurate imaging at various distances.

[0155] This embodiment also provides an electronic device 800, please refer to Figure 6 , Figure 6 A block diagram of the electronic device 800 is shown as an example. The electronic device includes a processor 810, a machine-readable storage medium 820 and a single-photon reconstruction device 830. The machine-readable storage medium 820 and the processor 810 are electrically connected to each other directly or indirectly to achieve data transmission or interaction. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines. The single-photon reconstruction device 830 includes a plurality of software function modules that can be stored in the machine-readable storage medium 820 in the form of software or firmware or solidified in the operating system (OS) of the single-photon reconstruction device 830. The processor 810 is used to execute executable modules stored in the machine-readable storage medium 820, such as software function modules and computer programs included in the single-photon reconstruction device 830.

[0156] The machine-readable storage medium 820 may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc. The machine-readable storage medium 820 is used to store a program, and the processor 810 executes the program after receiving an execution instruction.

[0157] The processor 810 may be an integrated circuit chip with signal processing capabilities. The above-mentioned processor 810 may be a general-purpose processor 810, including a central processing unit 810 (CPU), a network processor 810 (NP), etc.; it may also be a digital signal processor 810 (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor 810 may be a microprocessor 810 or the processor 810 may also be any conventional processor 810, etc.

[0158] Please refer to Figure 7 The present application also provides a single-photon reconstruction device 830, which includes a plurality of functional modules that can be stored in a machine-readable storage medium 820 in the form of software. Functionally, the single-photon reconstruction device 830 may include a construction module 831, an optimization module 832, a determination module 833, a secondary iterative solution module 834, and a generation module 835. Among them:

[0159] The construction module 831 is used to collect the real value of the photon, and simulate and generate a training set according to the real value of the photon, and the training set includes the simulated value of the photon.

[0160] In this embodiment, the construction module 831 can be used to execute Figure 1 As shown in step S110, for the detailed description of the construction module 831, reference may be made to the description of the step S110.

[0161] The optimization module 832 is used to input the photon simulation value into the denoising module 100 and the reconstruction module 200 constructed based on the attention mechanism in the deep learning network model for training to obtain a reconstructed point cloud.

[0162] In this embodiment, the optimization module 832 can be used to perform Figure 1 As shown in step S120, for the detailed description of the optimization module 832, please refer to the description of step S120.

[0163] The determination module 833 is used to input the reconstructed point cloud into the auxiliary training model of the deep learning network model to obtain the photon prediction value.

[0164] In this embodiment, the determination module 833 may be used to execute Figure 1As shown in step S130, for the specific description of the determination module 833, reference may be made to the description of the step S130.

[0165] The secondary iterative solution module 834 is used to calculate the loss of the deep learning network model and adjust the network parameters of the deep learning network model based on the predicted value of the photon and the actual value of the photon, so as to obtain a trained deep learning network model.

[0166] In this embodiment, the secondary iterative solution module 834 can be used to perform Figure 1 As shown in step S140, for the detailed description of the secondary iterative solution module 834, please refer to the description of step S140.

[0167] The generating module 835 is used to generate a depth image and an intensity image according to the ideal photon distribution of the target.

[0168] In this embodiment, the generating module 835 can be used to execute Figure 1 As shown in step S150, for the specific description of the generating module 835, reference may be made to the description of the step S150.

[0169] In summary, the present application provides a single-photon reconstruction method, device and electronic device. By constructing a photon counting model and iteratively solving the photon counting model, the photon distribution under the ideal target can be obtained. Compared with the existing technology, adaptive rapid imaging at close and long distances can be achieved, and the denoising effect can be effectively improved, and a more accurate three-dimensional spatial distribution of photons can be achieved statistically.

[0170] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0171] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A single photon reconstruction method, characterized in that: The method comprises: Acquire photon distribution observation values, and construct a photon counting model according to the photon distribution observation values; Using a negative Poisson maximum likelihood function as a loss function, optimizing the photon counting model, and determining an optimized constraint form; Determining a subproblem of the photon counting model according to the loss function and the optimized constraint form; Solve the sub-problem twice iteratively to obtain the photon distribution under the target ideal; A depth image and an intensity image are generated according to the ideal photon distribution of the target.

2. The single photon reconstruction method according to claim 1, characterized in that: The step of using the negative Poisson maximum likelihood function as a loss function to optimize the photon counting model and determine the optimized constraint form includes: According to the photon counting model, a probability function of photons is obtained; A negative Poisson maximum likelihood function is used as the loss function, and an optimized constraint form is obtained according to the probability function of the photon.

3. The single photon reconstruction method according to claim 2, characterized in that: The probability function of the photon is: Among them, e i represents the standard unit basis vector, y represents the observed value of the photon distribution, A represents the point spread function of the photon, f represents the photon distribution under the target ideal, and T represents the transpose; The loss function is: Among them, ε is a constant not equal to 0; The optimized constraint form is: minimizeΦ(f)≡F(f)+τpen(f) subjecttof≥0 Among them, τ represents a constant and pen represents a penalty term.

4. The single photon reconstruction method according to claim 1, characterized in that: The step of determining the sub-problems of the photon counting model according to the loss function and the optimized constraint form comprises: constructing an approximate expression according to the optimized constraint form, and using the difference as a penalty term of the approximate expression; Use the Taylor function to perform a quadratic expansion on the loss function to obtain a Taylor expansion; The subproblems of the photon counting model are obtained according to the approximate expression and the Taylor expansion.

5. The single photon reconstruction method according to claim 4, characterized in that: The approximate expression constructed according to the optimized constraint form is: subject to f≥0 Where f represents the photon distribution under the ideal target, k represents the number of iterations, represents the approximate value of the loss function; τ represents a constant; pen represents a penalty term; The Taylor expansion is: Among them, α k I represents the approximate value of the second-order derivative of the loss function, I represents the identity matrix, and F(f k ) represents the loss function, represents the first-order derivative of the loss function; The subproblems of the photon counting model are: subject to f≥0, in, 6. The single photon reconstruction method according to claim 1, characterized in that: The photon counting model is: y=Af Among them, y represents the observed value of photon distribution, A represents the point spread function of photons, and f represents the photon distribution under the ideal target.

7. The single photon reconstruction method according to claim 1, characterized in that: After the step of constructing a photon counting model according to the photon distribution observation value, the method further includes: Acquiring experimental equipment parameters, and calculating the point spread function of photons according to the experimental equipment parameters; The point spread function is simplified in the form of a Gaussian function to obtain the probability distribution of photons in space.

8. The single photon reconstruction method according to claim 7, characterized in that: The point spread function is: Among them, σ x , σ y , σ z represents the standard deviation, (x, y, z) represents the spatial coordinate position of the photon, and I0 represents a constant; The probability distribution of photons in space is: Among them, (a, b, c) represents the three-dimensional spatial position of the photon, t represents the time range of the photon, Spot represents the spot radius, fov represents the radius of the receiving field of view, and d represents the imaging distance.

9. A single photon reconstruction device, characterized in that: include: A construction module is used to obtain photon distribution observation values ​​and construct a photon counting model according to the photon distribution observation values; An optimization module, used to optimize the photon counting model using a negative Poisson maximum likelihood function as a loss function, and determine an optimized constraint form; A determination module, used to determine a sub-problem of the photon counting model according to the loss function and the optimized constraint form; A secondary iterative solution module is used to perform secondary iterative solutions to the sub-problems to obtain the photon distribution under the target ideal condition; A generation module is used to generate a depth image and an intensity image according to the ideal photon distribution of the target.

10. An electronic device, characterized in that: The electronic device includes a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by the processor, the method described in any one of claims 1 to 8 is implemented.

Images