Single-pixel three-dimensional non-invasive imaging method and device in random medium

Abstract

The invention provides a single-pixel three-dimensional non-invasive imaging method and device in a random medium, and relates to the technical field of computer imaging, and the method comprises the steps: dividing a to-be-imaged target through a virtual focusing scanning surface perpendicular to a light diffusion direction, and obtaining a plurality of vertical planes; according to the depth distance between each vertical plane in the random medium and the laser incident plane, parameters of the point spread functions of all the vertical planes are subjected to self-calibration processing; establishing a laser propagation model in a single-pixel confocal detection mode according to a self-calibration result; and calculating a to-be-imaged target by using the laser propagation model to obtain three-dimensional space distribution of the reflectivity of the to-be-imaged target. According to the scheme, non-intrusive and self-calibration efficient imaging can be carried out on the three-dimensional object immersed in the random medium.

Description

Technical Field

[0001] The present invention relates to the field of computer imaging technology, and in particular to a single-pixel three-dimensional non-invasive imaging method and device in a random medium. Background Art

[0002] The scattering effect of random media on light seriously restricts the imaging ability of optical detection and ranging systems through clouds, fog, rain, turbid water, sand and dust, biological tissues, etc. The method of inverting the target image by accurately simulating the light transmission process has the advantages of reliability and low computational complexity, and has important application potential in the field of imaging through random media.

[0003] In the related art, it is often difficult to achieve optical imaging of a three-dimensional target immersed in a random medium because light often experiences different degrees of diffusion when reaching the surface of an object at different depths in the random medium.

[0004] Based on this, there is an urgent need for a single-pixel three-dimensional non-invasive imaging method and device in a random medium to solve the above technical problems. Summary of the invention

[0005] The embodiments of the present invention provide a single-pixel three-dimensional non-invasive imaging method and device in a random medium, which can effectively realize imaging detection of a target immersed in the random medium.

[0006] In a first aspect, an embodiment of the present invention provides a single-pixel three-dimensional non-invasive imaging method in a random medium, comprising:

[0007] The target to be imaged is divided by using a virtual focusing scanning plane perpendicular to the light diffusion direction to obtain multiple vertical planes;

[0008] According to the depth distance between each vertical plane in the random medium and the laser incident plane, the parameters of the point spread function of all vertical planes are self-calibrated;

[0009] The laser propagation model in single-pixel confocal detection mode is established based on the self-calibration results;

[0010] The laser propagation model is used to calculate the target to be imaged, and the three-dimensional spatial distribution of the reflectivity of the target to be imaged is obtained.

[0011] In a second aspect, an embodiment of the present invention further provides a single-pixel three-dimensional non-invasive imaging device in a random medium, comprising:

[0012] A division module, used for dividing the target to be imaged by using a virtual focus scanning plane perpendicular to the light diffusion direction to obtain a plurality of vertical planes;

[0013] A calibration module is used for self-calibrating the parameters of the point spread function of all vertical planes according to the depth distance between each vertical plane in the random medium and the laser incident plane;

[0014] A modeling module, used to establish a laser propagation model in single-pixel confocal detection mode based on the self-calibration results;

[0015] The calculation module is used to calculate the target to be imaged by using the laser propagation model to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0016] In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of this specification is implemented.

[0017] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method described in any embodiment of this specification.

[0018] The embodiment of the present invention provides a single-pixel three-dimensional non-invasive imaging method and device in a random medium. First, the target to be imaged is divided into multiple vertical planes, and the parameters of the point spread function are self-calibrated according to the measured values ​​of the vertical planes. Then, a light propagation model in a single-pixel confocal detection mode is established according to the self-calibration results. Finally, the laser propagation model is used to calculate the target to be imaged after tomographic focusing analysis to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged. This method is a non-invasive, self-calibrated, and efficient imaging method. It can be used in a low-power laser imaging system with an average power of 0.18mW and a peak power of 750mW, and the number of detected photons is only a few hundred. Imaging in a weak light environment has great application potential in fields such as medicine, rescue, exploration, and autonomous driving where the target is usually immersed in a random medium. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 is a flow chart of a single-pixel three-dimensional non-invasive imaging method in a random medium provided by an embodiment of the present invention;

[0021] Figure 2 is a schematic diagram of the working of the tomofocusing method provided by one embodiment of the present invention;

[0022] Figure 3 is a schematic diagram of an imaging restoration result of the number "2" provided by an embodiment of the present invention;

[0023] Figure 4 is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0024] Figure 5 It is a structural diagram of a single-pixel three-dimensional non-invasive imaging device in a random medium provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0026] As mentioned earlier, existing imaging technologies can often only describe the diffusion behavior of light in a few specific random medium environments. It is difficult to describe the propagation process of light when actively detecting objects immersed in random media, resulting in poor imaging effects for such objects.

[0027] Based on this, the concept of the present invention is to use a single-pixel confocal detection method to determine the relationship between the reflectivity of the target plane and the measured light intensity of the target plane, and to restore the reflectivity through deconvolution. At the same time, the reflectivity of planes at different depths is restored layer by layer according to the tomographic focusing imaging method, thereby realizing imaging of objects immersed in random media.

[0028] The specific implementation of the above concept is described below.

[0029] Please refer to Figure 1 The embodiment of the present invention provides a single-pixel three-dimensional non-invasive imaging method in a random medium, the method comprising:

[0030] Step 100, dividing the target to be imaged by using a virtual focus scanning plane perpendicular to the light diffusion direction to obtain a plurality of vertical planes;

[0031] Step 102, according to the depth distance between each vertical plane in the random medium and the laser incident plane, the parameters of the point spread function of all vertical planes are self-calibrated;

[0032] Step 104, establishing a laser propagation model in a single-pixel confocal detection mode according to the self-calibration result;

[0033] Step 106 , using the laser propagation model to calculate the target to be imaged, to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0034] In the embodiment of the present invention, the target to be imaged is first divided into multiple vertical planes, the parameters of the point spread function are self-calibrated according to the measured values ​​of the vertical planes, and then the light propagation model in the single-pixel confocal detection mode is established according to the self-calibration results. Finally, the laser propagation model is used to calculate the target to be imaged after the tomographic focusing analysis to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged. This method is a non-invasive, self-calibrated, efficient imaging method that can be used in a low-power laser imaging system with an average power of 0.18mW and a peak power of 750mW, and imaging in a weak light environment with only hundreds of detected photons. It has great application potential in fields such as medicine, rescue, exploration, and autonomous driving where the target is usually immersed in a random medium.

[0035] Described below Figure 1 How the various steps are performed.

[0036] First, with respect to step 100 , the target to be imaged is divided using a virtual focus scanning plane perpendicular to the light diffusion direction to obtain a plurality of perpendicular planes.

[0037] Based on the imaging method for a planar object, an embodiment of the present invention proposes an imaging method for a volume target immersed in a random medium, which is called a tomographic focusing method.

[0038] This method is based on the fact that a three-dimensional object can be decomposed into a collection of surfaces in the z dimension, at different positions z m <z n The surface of the object, z = z m The signal reflected from the surface always lags behind z = z on average. n The signal reflected from the surface, the difference in the flight time of the straight-line photons on the two surfaces That is to say, if we can first restore the reflectivity distribution of the object's surface closest to the detector, then its reflectivity distribution can be separated from the detection signals of the rest of the object's surfaces.

[0039] Based on this fact, this embodiment uses a virtual focus scanning plane perpendicular to the light diffusion direction to divide the imaging target, and obtains multiple vertical planes for subsequent processing and restoration, such as Figure 2 shown.

[0040] Then, for step 102, according to the depth distance between each vertical plane in the random medium and the laser incident plane, the parameters of the point spread function of all vertical planes are self-calibrated.

[0041] The diffusion equation method is a commonly used method for simulating the transmission of light in a finite random medium by solving the point spread function of the diffusion of light in a random medium. However, the existing diffusion equation usually requires independent calibration of the parameters in the point spread function. Therefore, an embodiment of the present invention provides a parameter self-calibration method using a single-pixel confocal detection method, that is, emitting laser pulses and receiving detection signals at the same position.

[0042] From the above process, we can know that the target to be imaged immersed in the random medium can be divided into multiple vertical planes. When there is an object in the random medium, the light signal received by the detector at a spatial position point comes from two parts: one part is the backscattered light of the random medium itself, and the other part is the light reflected by the object, such as Figure 3 c is shown by the blue curve.

[0043] Therefore, in the embodiment of the present invention, the parameters of the point spread function are calibrated by the following method:

[0044] The starting time of the backscattered light of the random medium and the starting time of the reflected light of the vertical plane are calculated to obtain the depth distance of the laser in the random medium when it reaches the vertical plane;

[0045] According to the depth distance, a point spread function of a vertical plane located at the depth distance is determined.

[0046] In an embodiment of the present invention, the point spread function is determined by the following process: determining a photon count distribution value corresponding to the depth distance;

[0047] The photon count distribution value after logarithmization is derived in the time dimension to obtain the time value corresponding to the peak value of the vertical plane reflected light;

[0048] Calculating the scattering coefficient of the laser in the random medium according to the time value;

[0049] Based on the scattering coefficient, the point spread function is determined

[0050]

[0051] Where k = Dc is the diffusion coefficient; s is the scattering coefficient.

[0052] Specifically, the depth distance z0 of the object is estimated by the time difference between the starting time t0 of the random medium backscattered light and the starting time t1 of the vertical plane reflected light:

[0053]

[0054] Where c is the speed of light in the random medium; t2 is the time value corresponding to the peak value of the reflected light on the vertical plane; t0 is the starting time of the backscattered light in the random medium.

[0055] Furthermore, by solving the diffusion equation, the point spread function at the propagation distance z0 in an infinite random medium is used to determine In the form of:

[0056]

[0057] Where k = Dc is the diffusion coefficient, c is the speed of light in random media, s is the scattering coefficient, μ a is the absorption coefficient. In the experiment, we do not consider the absorption effect of random media, that is, let μ a =0.

[0058] The process of propagating the laser pulse in the random medium to the plane at z0 and returning to the incident point is approximately equivalent to the laser pulse propagating twice the distance z0 in the random medium, that is, the photon count distribution can be approximately expressed as Description. Take the logarithm and differentiate it with respect to the time dimension:

[0059]

[0060] The above formula is 0 at time t2 corresponding to the peak value of the plane reflected light.

[0061] at this time:

[0062]

[0063] The scattering coefficient can be obtained by the above formula:

[0064]

[0065] It can be understood that the self-calibration method of the parameters in the above point spread function can directly estimate the parameters according to the actual measured values ​​without the need for additional calibration measurement operations, thereby improving the timeliness and ease of operation of subsequent imaging.

[0066] With respect to step 104 , a laser propagation model in the single-pixel confocal detection mode is established according to the self-calibration result.

[0067] Multiple experiments have shown that when using a single-pixel confocal detection method, the target's light intensity measurement value can be written as the convolution between the target's reflectivity distribution and a convolution kernel, and the target reflectivity can be restored by only a single deconvolution.

[0068] Based on this, in the embodiment of the present invention, a laser propagation model in a single-pixel confocal detection mode is established by the following steps:

[0069] According to the point spread function and the initial light intensity distribution value of the random medium incident plane, the first light intensity distribution value when the laser propagates to the target vertical plane but does not interact with the target vertical plane is calculated.

[0070]

[0071] Among them, S(x, y, t) is the initial light intensity distribution value of the light source on the z=0 plane, that is, the incident plane of the random medium; is the point spread function;

[0072] According to the principle of mirror reflection, the first light intensity distribution value and the second light intensity distribution value after the laser propagates to the target vertical plane and interacts with the target vertical plane are established. The relationship equation is:

[0073]

[0074] Where, R(x,y,t)=ρ(x,y,z / c); R(x,y,t) is the reflectivity distribution value of the vertical plane in the incident direction;

[0075] According to the relationship equation and the light intensity measurement value of the single-pixel confocal detector, the laser propagation model is established as:

[0076]

[0077] in, is the light intensity measurement value of the z0 plane; is the convolution kernel of the z0 plane, which represents the point spread function of the laser propagating z0 distance in a random medium, being reflected by a reflection point, and then propagating back z0 distance.

[0078] The specific implementation process of the above steps is described below.

[0079] When the target is immersed in a random medium, the propagation path of the pulsed laser in the random medium can be described as follows: the laser pulse is incident on the random medium, diffuses in the random medium and propagates to the target plane, is reflected by the target plane and diffuses again in the random medium, and finally propagates to the detector to be detected, such as Figure 2 As shown in the dashed box in the upper left corner of the image.

[0080] In order to explore the relationship between the measured value τ(x, y, t) and the target reflectivity, and then restore the target reflectivity, we first consider the simplest detection scenario: the target is immersed in a random medium and is a two-dimensional plane object located in the z=z0 plane perpendicular to the laser propagation direction (set as the z direction). The pulsed laser is incident perpendicular to the z=0 plane. When the light propagates to the z=z0 plane but has not yet interacted with the target, the first light intensity distribution value of the light intensity in the z=z0 plane is:

[0081]

[0082] When using a pulse laser with extremely short temporal and spatial distribution, the light intensity distribution S(x, y, t) can be approximately considered as a pulse function δ(x, y, t). is the point spread function of the pulsed laser in the random medium with a propagation distance z0, is the convolution operator acting on the three dimensions x, y, and t.

[0083] The interaction between the target and the light is simplified to mirror reflection, and it is assumed that the light always illuminates the target vertically. The second light intensity distribution value on the z=z0 plane after the light interacts with the target is:

[0084]

[0085] Among them, ρ(x, y, t) is the spatiotemporal distribution of the target reflectivity in the z=z0 plane.

[0086] Assuming the target is a stationary target, ρ(x, y, t) is uniformly distributed in the t dimension. The effect is equivalent to Multiply by a constant, which is equivalent to convolving a delta pulse function. For a planar target perpendicular to the z direction, its reflectivity distribution ρ(x,y,z / c) in the z dimension is just a delta pulse function. In order to ensure the matching of the calculation operation dimensions, let R(x,y,t) = ρ(x,y,z / c), so the above formula can be rewritten as:

[0087]

[0088] in is a one-dimensional convolution along the t dimension.

[0089] After that, the light diffuses in the random medium in the direction opposite to the incident direction to the z=0 plane and is finally received by the detector. The light intensity measurement received by the detector is:

[0090]

[0091] because It is known that R(x,y,t) can be obtained by two deconvolutions (first three-dimensional deconvolution of x,y,t dimensions, then one-dimensional deconvolution of t dimension). However, there will be a lot of unknown noise in the actual detection process. Multiple deconvolution operations will amplify the interference of noise on the imaging results. Multiple deconvolution operations should be avoided.

[0092] In the single-pixel confocal detection mode adopted in this embodiment, the above formula can be expanded to:

[0093]

[0094] In this way, it can be converted into a convolution between R(x,y,t) and a convolution kernel:

[0095]

[0096] This is the random medium diffusion mode, that is, the laser propagation model. At this time, R(x, y, t) can be obtained through only one deconvolution process.

[0097] With respect to step 106 , the target to be imaged is calculated using the laser propagation model to obtain a three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0098] In an embodiment of the present invention, the three-dimensional spatial distribution of reflectivity is achieved by the following steps: calculating the convolution kernel of each vertical plane; substituting the convolution kernel into the laser propagation model to deconvolve the light intensity measurement value of each vertical plane; sampling the deconvolution result until all vertical planes of the target to be imaged are sampled, thereby obtaining the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0099] Specifically, after establishing the convolution relationship between the measured value modeling and the random medium diffusion pattern, the target reflectivity distribution and the random medium diffusion pattern are decoupled. Considering the existence of additive noise in the actual detection process, there are many decoupling methods, such as deconvolution filters, such as Wiener deconvolution filters, as well as optimization iterative methods and neural networks. The embodiment of the present invention uses the computationally simple Wiener deconvolution filtering technology to recover R(x, y, t):

[0100]

[0101] Where snr is a signal-to-noise ratio power spectrum. In the imaging scene through random media, the signal-to-noise ratio power spectrum is usually unknown and can be replaced by a constant. is Fourier transform; is the inverse Fourier transform.

[0102] Furthermore, after deconvolution is completed on each vertical plane, a plane with t=0 is selected to perform sampling processing on the deconvolution result.

[0103] In summary, the virtual focus scanning plane starts from the z=z0 plane, and the convolution kernel corresponding to the depth is Deconvolution is performed to restore the z=z0 plane of the object, the restored result of the z=z0 plane is frozen, and moved to the z=z1 plane, and the above steps are repeated to finally restore the reflectivity of the entire three-dimensional target.

[0104] The practicality of the above imaging method is verified by an experimental data.

[0105] like Figure 3 As shown, Figure 3 In the figure, a is the physical image of the number "2". b is the single photon counting image received by the detector. c is the distribution of photon counts received at a scanning point over time. d is the experimental scene image. e is the photon counting image of only the object reflected light obtained by subtracting the photon counting data when there is an object from the photon counting data when there is no object. f is the target reflectivity distribution image restored by the above method. All detection data are three-dimensional distributions (the lengths of the three dimensions of x, y, and t are 32, 32, and 128, respectively). For the convenience of display, all spatial distribution images are displayed with the maximum value in the t direction.

[0106] pass Figure 3 b, it can be seen that there are several obvious strong points in the upper left corner. This is because the laser irradiates the mirror reflection point on the random medium box almost vertically and returns along the original path. At this time, the detector receives not only scattered light but also directly returned mirror reflection light, resulting in a significantly higher photon count than other spatial positions. When the depth of the target is greater than several scattering free paths, the backscattered light of the random medium itself will be significantly stronger than the reflected light of the object, thus presenting the following Figure 3 The uniformly distributed image shown in b.

[0107] In order to pick out the light reflected by the target object we are concerned about, one method is to use a gating device to turn on the detector at the starting time t1 of the object's reflected light, or to collect all the photon count distributions and cut them off at the starting time t1 of the object's reflected light, and only retain the photon count data after time t1. This method will cause certain errors, because the photon count data after time t1 still contains a part of the backscattered light of the random medium itself. Another method is to pre-measure a set of photon count data of the random medium when there is no object as the background, or use the photon count data of the position where there is no object reflected light in the measurement value as the background, such as Figure 3 As shown in the orange curve in c, the background is subtracted from the photon counting data when an object is present to obtain a photon counting image containing only the light reflected by the object.

[0108] The photon counting image of the reflected light from the object is still severely distorted and the specific shape cannot be identified. After being processed by our restoration algorithm, the photon counting image becomes clear and sharp, and it is easy to identify the shape and size of the number "2", such as Figure 3 f.

[0109] like Figure 4 , Figure 5 As shown, the embodiment of the present invention provides a single-pixel three-dimensional non-invasive imaging device in a random medium. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 4 FIG. 1 is a hardware architecture diagram of an electronic device in which a single-pixel three-dimensional non-invasive imaging device in a random medium is provided in an embodiment of the present invention. Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the electronic device in which the device is located in the embodiment may also generally include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 5 As shown, as a device in a logical sense, the CPU of the electronic device in which it is located reads the corresponding computer program in the non-volatile memory into the memory and runs it. This embodiment provides a single-pixel three-dimensional non-invasive imaging device in a random medium, including:

[0110] A division module 500, used to divide the target to be imaged by using a virtual focus scanning plane perpendicular to the light diffusion direction to obtain a plurality of vertical planes;

[0111] A calibration module 502, configured to perform self-calibration processing on the parameters of the point spread functions of all vertical planes according to the depth distance between each vertical plane in the random medium and the laser incident plane;

[0112] A modeling module 504, for establishing a laser propagation model in a single-pixel confocal detection mode according to the self-calibration result;

[0113] The calculation module 506 is used to calculate the target to be imaged using the laser propagation model to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0114] In the embodiment of the present invention, when the calibration module 502 performs self-calibration processing on the parameters of the point spread functions of all vertical planes according to the depth distance between each vertical plane in the random medium and the laser incident plane, it is specifically used to perform the following operations: calculate the starting time of the backscattered light of the random medium and the starting time of the reflected light of the vertical plane to obtain the depth distance between the vertical plane and the laser incident plane; and determine the point spread function of the vertical plane at the depth distance according to the depth distance.

[0115] In the embodiment of the present invention, the depth distance is calculated by the following formula:

[0116]

[0117] Where c is the speed of light in the random medium; t2 is the time value corresponding to the peak value of the reflected light on the vertical plane; t0 is the starting time of the backscattered light in the random medium; and z0 is the depth distance.

[0118] In an embodiment of the present invention, according to the depth distance, determining the point spread function of a vertical plane at the depth distance includes: determining a photon count distribution value corresponding to the depth distance; performing a derivative calculation on the time dimension of the photon count distribution value after taking the logarithm to obtain a time value corresponding to a peak value of reflected light on the vertical plane; calculating a scattering coefficient of the laser in a random medium according to the time value; and determining a point spread function according to the scattering coefficient.

[0119]

[0120] Where k = Dc is the diffusion coefficient; s is the scattering coefficient.

[0121] In the embodiment of the present invention, when the modeling module 504 establishes the laser propagation model in the single-pixel confocal detection mode according to the self-calibration result, it is specifically used to perform the following operations: according to the point spread function and the initial light intensity distribution value of the random medium incident plane, calculate the first light intensity distribution value when the laser propagates to the target vertical plane but does not interact with the target vertical plane

[0122]

[0123] Among them, S(x,y,t) is the initial light intensity distribution value; is the point spread function;

[0124] According to the principle of mirror reflection, the first light intensity distribution value and the second light intensity distribution value after the laser propagates to the target vertical plane and interacts with the target vertical plane are established. The relationship equation is:

[0125]

[0126] Where, R(x,y,t)=ρ(x,y,z / c); R(x,y,t) is the reflectivity distribution value of the vertical plane in the incident direction;

[0127] According to the relationship equation and the light intensity measurement value of the single-pixel confocal detector, the laser propagation model is established as:

[0128]

[0129] in, is the light intensity measurement value of the z0 plane; is the convolution kernel in the z0 plane.

[0130] In the embodiment of the present invention, when the calculation module 506 uses the laser propagation model to calculate the target to be imaged and obtains the three-dimensional spatial distribution of the reflectivity of the target to be imaged, it is specifically used to perform the following operations: calculate the convolution kernel of each vertical plane; substitute the convolution kernel into the laser propagation model to deconvolve the light intensity measurement value of each vertical plane; sample the deconvolution result until all vertical planes of the target to be imaged are sampled to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

[0131] In the embodiment of the present invention, the light intensity measurement value is deconvoluted using the following formula:

[0132]

[0133] Where, snr is the signal-to-noise ratio power spectrum; is Fourier transform; is the inverse Fourier transform.

[0134] It is understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a single-pixel three-dimensional non-invasive imaging device in a random medium. In other embodiments of the present invention, a single-pixel three-dimensional non-invasive imaging device in a random medium may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0135] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0136] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a single-pixel three-dimensional non-invasive imaging method in a random medium in any embodiment of the present invention is implemented.

[0137] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a single-pixel three-dimensional non-invasive imaging method in a random medium in any embodiment of the present invention.

[0138] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0139] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0140] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0141] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0142] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0143] It should be noted that, in this article, relational terms such as first and second 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 statement "comprise a ..." do not exclude the presence of other identical factors in the process, method, article or device including the elements.

[0144] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0145] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A single-pixel three-dimensional non-invasive imaging method in a random medium, characterized in that: include: The target to be imaged is divided by using a virtual focusing scanning plane perpendicular to the light diffusion direction to obtain multiple vertical planes; According to the depth distance between each vertical plane in the random medium and the laser incident plane, the parameters of the point spread function of all vertical planes are self-calibrated; The laser propagation model in single-pixel confocal detection mode is established based on the self-calibration results; The laser propagation model is used to calculate the target to be imaged, and the three-dimensional spatial distribution of the reflectivity of the target to be imaged is obtained.

2. The method according to claim 1, characterized in that The method performs self-calibration processing on the parameters of the point spread functions of all vertical planes according to the depth distance between each vertical plane in the random medium and the laser incident plane, including: The starting time of the random medium backscattered light and the starting time of the vertical plane reflected light are calculated to obtain the depth distance between the vertical plane and the laser incident plane; According to the depth distance, a point spread function of a vertical plane located at the depth distance is determined.

3. The method according to claim 2, characterized in that The depth distance is calculated by the following formula: Where c is the speed of light in the random medium; t2 is the time value corresponding to the peak value of the reflected light on the vertical plane; t0 is the starting time of the backscattered light in the random medium; and z0 is the depth distance.

4. The method according to claim 3, characterized in that Determining a point spread function of a vertical plane located at the depth distance according to the depth distance includes: Determine a photon count distribution value corresponding to the depth distance; The photon count distribution value after logarithmization is derived in the time dimension to obtain the time value corresponding to the peak value of the vertical plane reflected light; Calculating the scattering coefficient of the laser in the random medium according to the time value; Based on the scattering coefficient, the point spread function is determined Where k = Dc is the diffusion coefficient; s is the scattering coefficient.

5. The method according to claim 1, characterized in that The method of establishing a laser propagation model in a single-pixel confocal detection mode according to the self-calibration result includes: According to the point spread function and the initial light intensity distribution value of the random medium incident plane, the first light intensity distribution value when the laser propagates to the target vertical plane but does not interact with the target vertical plane is calculated Among them, S(x,y,t) is the initial light intensity distribution value; is the point spread function; According to the principle of mirror reflection, the first light intensity distribution value and the second light intensity distribution value after the laser propagates to the target vertical plane and interacts with the target vertical plane are established. The relationship equation is: Where, R(x,y,t)=ρ(x,y,z / c); R(x,y,t) is the reflectivity distribution value of the vertical plane in the incident direction; According to the relationship equation and the light intensity measurement value of the single-pixel confocal detector, the laser propagation model is established as: in, is the light intensity measurement value of the z0 plane; is the convolution kernel in the z0 plane.

6. The method according to claim 5, characterized in that The method of calculating the target to be imaged by using the laser propagation model to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged includes: Calculate the convolution kernel for each vertical plane; Substituting the convolution kernel into the laser propagation model to deconvolve the light intensity measurement value of each vertical plane; The deconvolution result is sampled until all vertical planes of the target to be imaged are sampled, and the three-dimensional spatial distribution of the reflectivity of the target to be imaged is obtained.

7. The method according to claim 6, characterized in that The light intensity measurements are deconvoluted using the following formula: Where, snr is the signal-to-noise ratio power spectrum; is Fourier transform; is the inverse Fourier transform.

8. A single-pixel three-dimensional non-invasive imaging device in a random medium, characterized in that: include: A division module, used for dividing the target to be imaged by using a virtual focus scanning plane perpendicular to the light diffusion direction to obtain a plurality of vertical planes; A calibration module is used for self-calibrating the parameters of the point spread function of all vertical planes according to the depth distance between each vertical plane in the random medium and the laser incident plane; A modeling module, used to establish a laser propagation model in single-pixel confocal detection mode based on the self-calibration results; The calculation module is used to calculate the target to be imaged by using the laser propagation model to obtain the three-dimensional spatial distribution of the reflectivity of the target to be imaged.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images