Imaging method and imaging device

Abstract

The invention discloses an imaging method and an imaging device, and relates to the field of imaging, and the imaging method comprises the steps: enabling pulse laser to irradiate a to-be-measured target which rotates according to a time sequence, and obtaining the respective reflection laser of the to-be-measured target at a plurality of rotation angles; sequentially performing single-photon detection on the reflected laser to determine the flight time of a target photon in the reflected laser; taking the distribution result of the flight time obtained at the plurality of rotation angles as observation data; iteratively executing the following operations until the value of the optimization objective function meets a preset condition to obtain a three-dimensional image of the to-be-measured target: determining the probability that the observation data can be measured when the three-dimensional image of the to-be-measured target is the predicted three-dimensional image according to the rotation matrix of the to-be-measured target, the predicted three-dimensional image of the to-be-measured target and the observation data; determining a value of an optimization objective function according to the probability and the predicted three-dimensional image; and changing the predicted three-dimensional image according to the value of the optimized objective function.

Description

Technical Field

[0001] The present invention relates to the field of imaging, and in particular to an imaging method and an imaging device. Background Art

[0002] Long-distance imaging technology has important application value in target recognition, high-precision imaging and other fields, but it faces serious challenges in ultra-long-distance imaging (such as interstellar imaging or satellite-to-earth high-precision imaging). This is because traditional long-distance imaging technology is limited by the diffraction limit, and its lateral resolution decreases significantly with the increase of imaging distance, resulting in the inability to achieve high-resolution imaging in ultra-long-distance scenes.

[0003] Laser reflection tomography is a new method that breaks through the limitations of traditional long-distance imaging technology. It reconstructs the three-dimensional information of the detection target through one-dimensional distance detection of multi-angle pulsed lasers, getting rid of the dependence of the lateral resolution on the system aperture. This method only relies on the longitudinal resolution of the system, that is, the time resolution. However, the current laser reflection tomography technology usually uses traditional photoelectric detectors. The time resolution of such detectors is usually in the nanosecond range, which is difficult to meet the needs of high-resolution imaging at ultra-long distances. In addition, as the imaging distance increases, the optical signal carrying the information of the detection target will significantly attenuate, further limiting the imaging accuracy of the system. Summary of the invention

[0004] In view of this, the present invention provides an imaging method and an imaging device.

[0005] As a first aspect of the present invention, the imaging method comprises:

[0006] The pulse laser is irradiated onto the rotating target in a time sequence, and the reflected laser of the target at multiple rotation angles is obtained;

[0007] Perform single photon detection on the reflected laser to determine the distribution result of the flight time of the target photon in the reflected laser;

[0008] The distribution results of the flight time obtained under multiple rotation angles are counted as a photon counting time histogram; the photon counting time histogram constitutes the observation data;

[0009] The following operations are iteratively performed until the value of the optimization objective function satisfies a preset condition, and the predicted three-dimensional image of the target to be measured when the value of the optimization objective function satisfies the preset condition is determined as the three-dimensional image of the target to be measured:

[0010] Based on the Poisson likelihood model, according to the rotation matrix of the target to be measured, the predicted three-dimensional image of the target to be measured and the observed data, the probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined;

[0011] Determine the value of the optimization objective function according to the above probability and the predicted three-dimensional image;

[0012] According to the value of the optimization objective function, the predicted three-dimensional image is changed.

[0013] According to an embodiment of the present invention, determining the value of the optimization objective function according to the above-mentioned probability and the predicted three-dimensional image includes:

[0014] Taking the negative logarithm of the above probability, a first processing result is obtained;

[0015] Performing total variation regularization processing on the observed data to obtain a second processing result;

[0016] The value of the optimization objective function is determined according to the sum of the first processing result and the second processing result.

[0017] According to an embodiment of the present invention, changing the predicted three-dimensional image optimization objective function of the target to be measured includes:

[0018] Calculate the gradient of the optimization objective function and change the predicted 3D image in the direction of gradient descent.

[0019] According to an embodiment of the present invention, based on a Poisson likelihood model, according to a rotation matrix of the target to be measured, a predicted three-dimensional image of the target to be measured, and the observed data, determining the probability that the observed data can be measured when the three-dimensional image of the target to be measured is a predicted three-dimensional image includes:

[0020] Determine ideal distribution results of the number of photons of the predicted three-dimensional image at multiple rotation angles according to the rotation matrix of the target to be measured and the predicted three-dimensional image of the target to be measured;

[0021] Based on the Poisson likelihood model, according to the ideal distribution results of the number of photons at multiple rotation angles and the observed data, the probability of measuring the observed data when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined.

[0022] According to an embodiment of the present invention, the preset condition is that the change in the value of the optimization objective function is less than a preset value.

[0023] As a second aspect of the present invention, an imaging device is also provided for implementing the above imaging method, the imaging device comprising:

[0024] The laser module is adapted to emit pulsed lasers, so that the pulsed lasers are irradiated onto the rotating target in a time sequence, so as to obtain respective reflected lasers of the target at multiple rotation angles;

[0025] Single photon detector, suitable for single photon detection of reflected laser light;

[0026] A time-to-digital converter is used to determine the distribution result of the flight time of the target photons detected by the single-photon detector in the reflected laser, and the distribution results of the flight time obtained at multiple rotation angles are statistically analyzed as a photon counting time histogram; the photon counting time histogram constitutes the observation data;

[0027] The processing module is adapted to iteratively perform the following operations until the value of the optimization objective function satisfies a preset condition, and the predicted three-dimensional image of the target to be measured when the value of the optimization objective function satisfies the preset condition is determined as the three-dimensional image of the target to be measured:

[0028] Based on the Poisson likelihood model, the probability of measuring the observed data when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined according to the rotation matrix of the target to be measured, the predicted three-dimensional image of the target to be measured and the observed data;

[0029] Determining a value of an optimization objective function based on the probability and the predicted three-dimensional image;

[0030] According to the value of the optimization objective function, the predicted three-dimensional image is changed.

[0031] According to an embodiment of the present invention, the imaging device further includes:

[0032] The transmission module is suitable for collimating the pulse laser output by the laser module so that the collimated pulse laser can be irradiated onto the target to be measured; and is suitable for collimating the reflected laser so that the collimated reflected laser can enter the single-photon detector.

[0033] The transport module includes:

[0034] A circulator, adapted to transmit pulsed laser light and to transmit received reflected laser light to a single photon detector;

[0035] The collimator is suitable for collimating the pulse laser from the circulator and irradiating the collimated pulse laser to the target to be measured, and is suitable for collimating the reflected laser from the target to be measured and transmitting it to the circulator.

[0036] According to an embodiment of the present invention, the laser module comprises:

[0037] A signal generator, suitable for sending control signals in a time sequence;

[0038] A laser, adapted to generate pulsed laser light in response to a control signal;

[0039] According to an embodiment of the present invention, the imaging device further includes:

[0040] The first optical switch allows the pulsed laser to be incident on the target to be measured when it is turned on, and prohibits the reflected laser from entering the laser module when it is turned off;

[0041] a second optical switch, which transmits the reflected laser light to the single-photon detector when turned on, and prohibits the pulsed laser light from entering the single-photon detector when turned off;

[0042] Wherein, in the emission phase of the pulsed laser, the first optical switch is in an open state, and the second optical switch is in a closed state; in the reception phase of the reflected laser, the first optical switch is in a closed state, and the second optical switch is in an open state;

[0043] The signal generator is further adapted to control the opening and closing of the first optical switch and the opening and closing of the second optical switch.

[0044] According to an embodiment of the present invention, the time-to-digital converter is suitable for determining the emission time of the pulsed laser, and determining the time when the single-photon detector receives the target photon of the reflected laser according to the electrical signal emitted by the single-photon detector after the pulsed laser is emitted, thereby calculating the flight time of the target photon.

[0045] According to an embodiment of the present invention, the signal generator is further adapted to send out a timing signal while sending out the control signal, so that the time-to-digital converter determines the emission time of each pulse laser according to the timing signal.

[0046] The embodiment of the present invention utilizes single-photon detection technology to obtain the distribution of flight time of target photons at different rotation angles. To address the Poisson noise problem in single-photon detection, the embodiment of the present invention utilizes a Poisson likelihood model and an optimization objective function to achieve the fusion of observed data and predicted three-dimensional images. By iteratively optimizing the predicted three-dimensional image, the image is gradually approximated to the three-dimensional image of the target to be measured, thereby organically combining single-photon detection and laser reflection tomography technology to achieve high-precision three-dimensional imaging of the rotating target to be measured.

[0047] Single-photon detection of the reflected laser and detection of the flight time of the target photon in the reflected laser simultaneously realize the combination of single-photon detection technology and time gating technology, and the time resolution that can be achieved is in the order of picoseconds, overcoming the problem of low time resolution when detecting the reflected laser in the existing laser reflection tomography technology, and achieving higher imaging accuracy. In addition, single-photon detection has high sensitivity, and even if the reflected light signal decays to an extremely low level (even close to the single photon level) after long-distance transmission, these weak light signals can still be effectively captured by single-photon detection of the reflected laser. Therefore, single-photon detection of the reflected laser in sequence can avoid the problem of poor imaging accuracy caused by long-distance transmission of the reflected light signal in the existing laser reflection tomography technology.

[0048] According to an embodiment of the present invention, in ultra-long-distance imaging, due to the long transmission distance of the reflected laser, the number of reflected photons that can be detected is very limited and obeys the Poisson distribution. The Poisson likelihood model can accurately describe the statistical characteristics of the data under such low photon count conditions, thereby more accurately reflecting the relationship between the observed data and the real target to be measured. The Poisson likelihood model can fully consider the Poisson noise characteristics in the observed data, thereby effectively suppressing the interference of Poisson noise during the processing process. Total variation (TV) regularization constrains the reconstruction result by introducing the sparsity prior of the image, that is, assuming that the image is sparse in the gradient domain, which can make the reconstructed image smoother and help suppress artifacts and noise during the imaging process, so that the image has higher clarity and accuracy. Convex optimization processing (i.e., repeated iterative process) has the uniqueness of the global optimal solution, which means that the global optimal reconstruction result can be obtained by solving the optimization objective function through the convex optimization method. The Poisson likelihood modeling of the observed data, the total variation regularization of the predicted three-dimensional image, and the convex optimization of the optimization objective function are to achieve noise resistance, suppress artifacts, maintain edge features, ensure the global optimal solution, and improve algorithm stability and efficiency in ultra-long-distance single-photon reflection tomography. These processing methods work together in the imaging process to help achieve high-precision ultra-long-distance three-dimensional imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of an imaging method provided according to an embodiment of the present invention is shown;

[0050] Figure 2 A schematic diagram of an imaging device provided according to an embodiment of the present invention is shown;

[0051] Figure 3 A comparison diagram of multiple images obtained by imaging a target to be measured using multiple methods provided in an embodiment of the present invention is shown.

[0052] Description of Reference Numerals

[0053] 1 laser module;

[0054] 11. Signal generator;

[0055] 12 lasers;

[0056] 2 single photon detectors;

[0057] 3Time to digital converter;

[0058] 4 transmission module;

[0059] 41 Circulator;

[0060] 42 collimator;

[0061] 5 first optical switch;

[0062] 6 Second optical switch. DETAILED DESCRIPTION

[0063] In the process of realizing the present invention, it was found that single-photon detection technology has gradually attracted attention in the field of high-precision imaging. Single-photon detection technology is usually used in combination with time-correlated single photon counting (TCSPC) technology. When the two are combined, it not only has single-photon level detection sensitivity, but also can achieve picosecond (ps) level time resolution. However, there are still some problems in directly applying single-photon detection technology to laser reflection tomography technology. For example, since the signal of single-photon detection has statistical characteristics, the image reconstruction process using the signal of single-photon detection is easily affected by Poisson noise. In this case, the traditional image reconstruction algorithm based on back projection is prone to artifacts, which cannot meet the requirements of high-precision imaging.

[0064] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0065] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0066] All terms used herein, including technical and scientific terms, have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0067] When using expressions such as "at least one of A, B, and C, etc.", it should generally be interpreted as the meaning of the expression generally understood by those skilled in the art. For example, "a system having at least one of A, B, and C" should include but not be limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc. When using expressions such as "at least one of A, B, or C, etc.", it should generally be interpreted as the meaning of the expression generally understood by those skilled in the art. For example, "a system having at least one of A, B, or C" should include but not be limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.

[0068] It should also be noted that the directional terms mentioned in the embodiments, such as "upper", "lower", "front", "back", "left", "right", etc., are only reference directions of the drawings and are not intended to limit the scope of protection of the present invention. Throughout the drawings, the same elements are represented by the same or similar reference numerals. Conventional structures or configurations will be omitted when they may cause confusion in the understanding of the present invention.

[0069] In the embodiments of the present invention, the collection, updating, analysis, processing, use, transmission, provision, invention, storage and other aspects of the data involved (for example, including but not limited to user personal information) are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information and network security.

[0070] Figure 1 A flow chart of an imaging method provided according to an embodiment of the present invention is shown.

[0071] like Figure 1 As shown, the imaging method includes operations S1 to S7.

[0072] In operation S1, pulse laser is irradiated onto the rotating target in time sequence, and respective reflected lasers of the target at multiple rotation angles are obtained. The pulse laser is reflected by the target after being incident on it, and the target may be, for example, a satellite or the earth.

[0073] In operation S2, single photon detection is performed on the reflected laser light in sequence to determine the distribution result of the flight time of the target photons in the reflected laser light. The target photons are the photons detected in the reflected laser light.

[0074] In operation S3, the distribution results of the flight time obtained at multiple rotation angles are counted as a photon counting time histogram; wherein the photon counting time histogram constitutes the observation data.

[0075] In operation S4, the following operations are iteratively performed until the value of the optimization objective function satisfies a preset condition, and the predicted three-dimensional image of the target to be measured when the value of the optimization objective function satisfies the preset condition is determined as the three-dimensional image of the target to be measured:

[0076] In operation S5, based on the Poisson likelihood model, the probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined according to the rotation matrix of the target to be measured, the predicted three-dimensional image of the target to be measured, and the observed data;

[0077] In operation S6, determining a value of an optimization objective function according to the above probability and the predicted three-dimensional image;

[0078] In operation S7, the predicted three-dimensional image is modified according to the value of the optimization objective function.

[0079] The irradiation range of the pulse laser in the embodiment of the present invention can cover a part of the target to be measured (i.e., the surface facing the pulse laser), and the irradiated part is the detection area of ​​the pulse laser. Therefore, when the target to be measured rotates, the detection area changes, but there is overlap between different detection areas. As the target to be measured rotates, each position on the target to be measured can be detected by the pulse laser at different rotation angles. The observation data in the embodiment of the present invention includes the detection information of any position of the target to be measured at different rotation angles.

[0080] According to an embodiment of the present invention, for each rotation angle of the target to be measured, multiple pulse lasers are used to irradiate the target to be measured at that rotation angle. This is because single-photon detection has high sensitivity and high time resolution, and multiple measurements are required to improve the statistical accuracy of the data.

[0081] The principle of laser reflection tomography is to analyze the reflection characteristics of laser on the surface or inside of an object, combine multi-angle measurement and computer algorithm reconstruction, and achieve high-resolution imaging of the object. The embodiment of the present invention uses single-photon detection technology to obtain the distribution of the flight time of target photons at different rotation angles. In view of the Poisson noise problem in single-photon detection, the embodiment of the present invention uses the Poisson likelihood model and the optimization objective function to achieve the fusion of the observed data and the predicted three-dimensional image. By iteratively optimizing the predicted three-dimensional image, it gradually approaches the three-dimensional image of the target to be measured, thereby organically combining single-photon detection and laser reflection tomography technology to achieve high-precision three-dimensional imaging of the rotating target to be measured.

[0082] According to an embodiment of the present invention, single-photon detection is performed on the reflected laser, and the flight time of the target photon in the reflected laser is detected at the same time, so that the single-photon detection technology and the time gating technology are combined, and the time resolution that can be achieved is in the order of picoseconds, which overcomes the problem of low time resolution when detecting the reflected laser in the existing laser reflection tomography technology, and can achieve higher imaging accuracy. In addition, single-photon detection has a high sensitivity. Even if the reflected light signal decays to an extremely low level (even close to the single photon level) after long-distance transmission, these weak light signals can still be effectively captured by performing single-photon detection on the reflected laser. Therefore, performing single-photon detection on the reflected laser in sequence can avoid the problem of poor imaging accuracy caused by long-distance transmission of the reflected light signal in the existing laser reflection tomography technology.

[0083] According to an embodiment of the present invention, operations S4 to S7 are a Poisson noise convex optimization reconstruction algorithm provided by an embodiment of the present invention. In ultra-long-distance imaging, due to the long transmission distance of the reflected laser, the number of reflected photons that can be detected is very limited and obeys the Poisson distribution. The Poisson likelihood model can accurately describe the statistical characteristics of the data under such low photon count conditions, thereby more accurately reflecting the relationship between the observed data and the real target to be measured. The Poisson likelihood model can fully consider the Poisson noise characteristics in the observed data, thereby effectively suppressing the interference of Poisson noise during the processing process. Total variation (TV) regularization constrains the reconstruction result by introducing the sparsity prior of the image, that is, assuming that the image is sparse in the gradient domain, which can make the reconstructed image smoother, and help suppress artifacts and noise during the imaging process, so as to improve the clarity and accuracy of the image. The Poisson noise convex optimization reconstruction algorithm has the uniqueness of the global optimal solution, which means that the global optimal reconstruction result can be obtained by solving the optimization objective function through the Poisson noise convex optimization reconstruction algorithm. The use of a Poisson likelihood model to process the observed data, the total variation regularization processing of the predicted three-dimensional image, and the convex optimization processing of the optimization objective function are to have anti-noise performance, suppress artifacts, maintain edge features, ensure the global optimal solution, and improve algorithm stability and efficiency in ultra-long-distance single-photon reflection tomography. These processing methods work together in the imaging process, which helps to achieve high-precision ultra-long-distance three-dimensional imaging. According to an embodiment of the present invention, based on the Poisson likelihood model, the probability of being able to measure the observed data when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined according to the rotation matrix of the target to be measured, the predicted three-dimensional image of the target to be measured, and the observed data, including: operations S51 to S52.

[0084] In operation S51, the ideal distribution results of the number of photons of the predicted three-dimensional image at multiple rotation angles are determined according to the rotation matrix of the target to be measured and the predicted three-dimensional image of the target to be measured. When the irradiation range of the pulsed laser can cover the surface of the target to be measured facing the pulsed laser, the ideal distribution results of the number of photons at multiple rotation angles are It is expressed as formula (1).

[0085] (1).

[0086] in, is the rotation matrix of the target to be measured, representing the transformation of the target to be measured at different rotation angles; is the predicted three-dimensional image of the target to be measured, Represents the three-dimensional spatial coordinates of different positions of the target to be measured; is the lateral distribution of pulsed laser (generally Gaussian light); is the time domain waveform of the pulsed laser, is the speed of light, represents the integration space, which includes all areas of the target to be measured that are irradiated by the pulse laser. Represents the predicted 3D image After the rotation matrix The result after transformation.

[0087] In operation S52, based on the Poisson likelihood model, according to the ideal distribution results of the number of photons at multiple rotation angles and the observed data, the probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined, as shown in formula (2).

[0088] (2).

[0089] in, It indicates the probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image. is the background noise, is the quantum efficiency of the detector .

[0090] Based on the maximum likelihood estimation, the optimization objective function is established. The optimization objective function is expressed as formula (3).

[0091] (3).

[0092] in, represents the optimization objective function, The three-dimensional image representing the target to be measured is the predicted three-dimensional image When The probability of Indicates different rotation angles of the target to be measured. Indicates the flight time, Represents the three-dimensional coordinates of different positions of the target to be measured, Represents the total variation regularization term of the predicted 3D image, which is used to preserve the edge of the predicted 3D image and avoid edge blur. TV represents the total variation, represents the regularization parameter.

[0093] According to formula (3), in operation S6, the value of the optimization objective function is determined according to the above probability and the predicted three-dimensional image, including operations S61 to S63.

[0094] In operation S61, the probability is negatively logarithmized to obtain a first processing result.

[0095] In operation S62, total variation regularization processing is performed on the observed data to obtain a second processing result;

[0096] In operation S63, a value of an optimization objective function is determined according to the sum of the first processing result and the second processing result.

[0097] According to an embodiment of the present invention, when changing the predicted three-dimensional image, non-negative constraints and total variation regularization are applied to ensure that the updated predicted three-dimensional image satisfies the non-negative constraints (i.e., the updated predicted three-dimensional image ≥0). If the updated predicted 3D image If there is a negative value in , it is set to zero. In addition, the updated predicted 3D image Total variation regularization is applied to further suppress noise and maintain the sparsity of the image.

[0098] Figure 2 A schematic diagram of an imaging device according to an embodiment of the present invention is shown.

[0099] like Figure 2 As shown, the imaging device includes: a laser module 1, a single photon detector 2, a time-to-digital converter 3 and a processing module (not shown in the figure).

[0100] The laser module 1 is suitable for emitting a pulsed laser, so that the pulsed laser is irradiated on the rotating target to be measured in a time sequence, so as to obtain the reflected laser of the target to be measured at multiple rotation angles. The single-photon detector 2 is suitable for receiving the reflected laser and performing single-photon detection on the reflected laser. The time-to-digital converter 3 is suitable for determining the flight time of all target photons detected by the single-photon detector 2 in the reflected laser, and statistically analyzing the distribution results of all flight times obtained at multiple rotation angles as a photon counting time histogram, that is, the photon counting time histogram is the statistical distribution of all flight times obtained at multiple rotation angles, and the photon counting time histogram constitutes the observation data. The processing module is suitable for iteratively performing the following operations until the value of the optimization objective function meets the preset conditions, and the predicted three-dimensional image of the target to be measured when the value of the optimization objective function meets the preset conditions is determined as the three-dimensional image of the target to be measured: based on the Poisson likelihood model, according to the rotation matrix of the target to be measured, the predicted three-dimensional image of the target to be measured and the observation data, the probability of being able to measure the observation data when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined; the value of the optimization objective function is determined according to the above probability and the predicted three-dimensional image.

[0101] According to an embodiment of the present invention, the single photon detector 2 and the time digital converter 3 are combined, and the problems of the traditional technology in time resolution and signal attenuation are solved through the sensitivity of the single photon detector 2 and the picosecond time resolution of the time digital converter 3, thereby improving the ultra-long-distance imaging capability. In addition, the present invention combines the Poisson noise convex optimization reconstruction algorithm to effectively suppress artifacts, so that the obtained three-dimensional image has high accuracy and robustness.

[0102] According to an embodiment of the present invention, the laser module 1 comprises: a signal generator 11 and a laser 12. The signal generator 11 is adapted to send out a control signal in a time sequence. The laser 12 is adapted to generate a pulsed laser according to the control signal.

[0103] According to an embodiment of the present invention, the imaging device further comprises: a transmission module 4. The transmission module 4 is adapted to collimate the pulsed laser so that the pulsed laser is incident on the target to be measured, and is adapted to collimate the reflected laser so that the reflected laser enters the single photon detector.

[0104] According to an embodiment of the present invention, the imaging device adopts a coaxial transceiver design, that is, the pulse laser is emitted from the transmission module 4, and the reflected laser is received by the transmission module 4. This design can greatly reduce the system volume, improve the integration, and meet the portability requirements.

[0105] According to an embodiment of the present invention, the transmission module 4 includes: a circulator 41 and a collimator 42 .

[0106] The circulator 41 is suitable for transmitting pulsed laser light, and for transmitting the received reflected laser light to the single-photon detector 2. The collimator 42 is suitable for collimating the pulsed laser light from the circulator, and allowing the collimated pulsed laser light to be incident on the target to be measured, and for transmitting the reflected laser light from the target to be measured to the circulator 41 after collimation.

[0107] According to an embodiment of the present invention, the signal generator 11 is further configured to send out a timing signal while sending out a control signal, so that the time-to-digital converter determines the emission time of the pulse laser according to the timing signal.

[0108] According to an embodiment of the present invention, the imaging device further includes: a first optical switch 5 and a second optical switch 6. The first optical switch 5 allows the pulsed laser to be incident on the target to be measured when it is turned on, and prohibits the reflected laser from entering the laser module 1 when it is turned off. The second optical switch 6 is used to transmit the reflected laser to the single-photon detector 2 when it is turned on, and prohibits the pulsed laser from entering the single-photon detector 2 when it is turned off. The first optical switch 5 and the second optical switch 6 are configured to be turned on alternately. In the pulsed laser emission stage, the first optical switch 5 is in the on state and the second optical switch 6 is in the off state, so that the pulsed laser is smoothly emitted. In the single-photon detector 2 receiving stage of the reflected laser (from the target to be measured reflecting the pulsed laser to obtain the reflected laser to the reflected laser being received by the single-photon detector 2), the first optical switch 5 is turned off and the second optical switch 6 is turned on, so that the reflected laser can be detected by the single-photon detector 2.

[0109] The signal generator 11 is also adapted to send a timing signal while sending a control signal, so that the time digital converter 3 determines the emission time of the laser pulse laser according to the timing signal. The signal generator 11 is also adapted to control the opening and closing of the first optical switch 5 and the opening and closing of the second optical switch 6. While the signal generator sends a control signal, the signal generator 11 also generates a high-level signal to the first optical switch 5, enabling the first optical switch 5 so that the pulse laser can be smoothly emitted through the collimator. In the reflected laser receiving stage, the signal generator 11 generates a high-level signal to the second optical switch 6, enabling the second switch 6, so that the echo energy (reflected laser) can be smoothly detected by the single photon detector 2, and at the same time generates a synchronization signal to the time digital converter for signal statistics.

[0110] The embodiment of the present invention adopts a time-sharing control of transmission and reception for the two optical switches, that is, the imaging device of the embodiment of the present invention uses half of the time for emitting pulsed lasers (the first optical switch is turned on and the second optical switch is turned off), and the other half of the time for receiving reflected signals (the second optical switch is turned on and the first optical switch is turned off). This design avoids noise interference caused by local optical signal reflection.

[0111] The single photon detector according to the embodiment of the present invention may be, for example, a miniaturized free-running InGaAs / InP single photon avalanche diode.

[0112] The following combination Figure 2 , and lists specific embodiments to explain in detail the imaging method provided by the embodiments of the present invention.

[0113] According to an embodiment of the present invention, taking the imaging distance of 6 km as an example, the target to be measured is fixed on a turntable 6 km away from the imaging device, and the turntable rotates 1 degree each time to make the target to be measured rotate. The laser module 1 and the first optical switch 5 are used to incident the pulse laser on the target to be measured in time sequence, and multiple pulse lasers are incident on the target to be measured after passing through the transmission module 4 and then reflected by the target to be measured. The reflected laser is sequentially transmitted to the single photon detector 2 through the transmission module 4 and the second optical switch 6 to complete the 360-degree all-round data collection of the target to be measured. According to the target distance between the imaging device and the target to be measured, the time when the pulse laser reaches the target to be measured can be estimated. For example, when the target distance is 6 km, it can be estimated that the time when the pulse laser reaches the target to be measured is 40 microseconds, so the pulse laser emission time and the reflected laser reception time can be set to 40 microseconds. According to the estimated time, the signal generator 11 is used to accurately control the opening and closing time of the first optical switch 5 and the second optical switch 6 to ensure the accurate collection of the reflected laser. The single photon detector 2 can be set to have a time interval (bin) width of 1 picosecond to obtain high-precision time resolution. Each acquisition lasts for 1 second, and the pulse laser repetition rate is 1MHz, ensuring real-time data update and high sampling rate. The time-to-digital converter is used to determine the flight time of the target photons in the reflected laser and count them into a photon counting time histogram. The processing module uses the Poisson noise convex optimization reconstruction algorithm to process the observation data obtained by the reflected laser at multiple rotation angles, and achieves high-precision three-dimensional reconstruction of the target to be measured. Finally, after algorithm optimization and reconstruction, a three-dimensional image of the target to be measured is obtained. The three-dimensional image has a three-dimensional highlighting effect, can accurately display the spatial contour of the target to be measured, and provides high-quality imaging data for the identification and detection of the target to be measured.

[0114] Figure 3 The comparison results of multiple images obtained by imaging the target to be measured using various methods are shown.

[0115] like Figure 3 As shown, part (a) is an image obtained by taking a close-up shot of the target to be measured using a visible light camera, and this image is the imaging target to be achieved by the imaging method of the embodiment of the present invention. Part (b) is a schematic diagram of the observation data obtained according to the embodiment of the present invention, part (c) is a three-dimensional image of the target to be measured obtained by using a traditional imaging method (traditional laser reflection tomography algorithm), and part (d) is a three-dimensional image of the target to be measured obtained by using the imaging method provided by the embodiment of the present invention. The scales of parts (a) and (d) are both 10 cm. It can be seen from parts (c) and (d) that the imaging method provided by the embodiment of the present invention can accurately obtain a three-dimensional image of the target to be measured, thereby realizing the reconstruction of the target to be measured.

[0116] The imaging method provided by the embodiment of the present invention exhibits significant noise resistance when processing photon counting data under a Poisson noise background, can achieve high-precision imaging under complex noise conditions, and adapt to parameter adjustment under different signal-to-noise ratio conditions.

[0117] The imaging method provided by the embodiment of the present invention utilizes single-photon detection technology and combines it with a Poisson noise convex optimization reconstruction algorithm, which can effectively overcome the attenuation of reflected signals and the influence of noise during ultra-long-distance imaging, and significantly improve the resolution imaging capability at ultra-long distances.

[0118] The imaging device provided in the embodiment of the present invention adopts a coaxial transceiver single-photon laser reflection tomography system configuration, which not only reduces the system volume, but also improves the system integration and reduces the hardware complexity and cost.

[0119] The embodiments of the present invention are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the present invention. Although each embodiment is described above, it does not mean that the measures in each embodiment cannot be used in combination. The scope of the present invention is defined by the attached claims and their equivalents. Without departing from the scope of the present invention, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

Claims

1. An imaging method, characterized in that: include: The pulsed laser is irradiated onto the rotating target in a time sequence, and the reflected lasers of the target at multiple rotation angles are obtained; Performing single-photon detection on the reflected laser to determine a distribution result of the flight time of target photons in the reflected laser; The distribution results of the flight time obtained under the multiple rotation angles are counted as a photon counting time histogram; the photon counting time histogram constitutes the observation data; Iteratively perform the following operations until the value of the optimization objective function meets a preset condition, and determine the predicted three-dimensional image of the target to be measured when the value of the optimization objective function meets the preset condition as the three-dimensional image of the target to be measured: Based on a Poisson likelihood model, according to a rotation matrix of the target to be measured, a predicted three-dimensional image of the target to be measured, and the observed data, determining a probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image; Determining a value of the optimization objective function according to the probability and the predicted three-dimensional image; The predicted three-dimensional image is modified according to the value of the optimization objective function.

2. The imaging method according to claim 1, characterized in that: Determining the value of the optimization objective function according to the probability and the predicted three-dimensional image includes: Taking a negative logarithm of the probability to obtain a first processing result; Performing total variation regularization processing on the observed data to obtain a second processing result; The value of the optimization objective function is determined according to the sum of the first processing result and the second processing result.

3. The imaging method according to claim 1, wherein: Changing the predicted three-dimensional image comprises: The gradient of the optimization objective function is calculated, and the predicted three-dimensional image is modified along the direction of gradient descent.

4. The imaging method according to claim 1, characterized in that: Based on a Poisson likelihood model, according to a rotation matrix of the target to be measured, a predicted three-dimensional image of the target to be measured, and the observed data, determining a probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image, including: Determine, according to the rotation matrix of the target to be measured and the predicted three-dimensional image of the target to be measured, the ideal distribution results of the number of photons of the predicted three-dimensional image at the multiple rotation angles; Based on the Poisson likelihood model, according to the ideal distribution results of the number of photons at each of the multiple rotation angles and the observed data, the probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image is determined.

5. The imaging method according to claim 2, characterized in that: The preset condition is that the change in the value of the optimization objective function is less than a preset value.

6. An imaging device for implementing the imaging method according to any one of claims 1 to 5, characterized in that: The imaging device comprises: The laser module is adapted to emit pulsed laser light, so that the pulsed laser light is irradiated onto the rotating target to be measured in a time sequence, so as to obtain respective reflected laser light of the target to be measured at multiple rotation angles; A single photon detector, adapted to perform single photon detection on the reflected laser; A time-to-digital converter, adapted to determine the distribution result of the flight time of the target photons detected by the single-photon detector in the reflected laser, and to count the distribution results of the flight time obtained at multiple rotation angles as a photon counting time histogram; the photon counting time histogram constitutes the observation data; The processing module is adapted to iteratively perform the following operations until the value of the optimization objective function satisfies a preset condition, and determine the predicted three-dimensional image of the target to be measured when the value of the optimization objective function satisfies the preset condition as the three-dimensional image of the target to be measured: Based on a Poisson likelihood model, according to a rotation matrix of the target to be measured, a predicted three-dimensional image of the target to be measured, and the observed data, determining a probability that the observed data can be measured when the three-dimensional image of the target to be measured is the predicted three-dimensional image; The value of the optimization objective function is determined according to the probability and the predicted three-dimensional image; and the predicted three-dimensional image is modified according to the value of the optimization objective function.

7. The imaging device according to claim 6, characterized in that The imaging device further comprises: A transmission module, adapted to collimate the pulsed laser output by the laser module so that the collimated pulsed laser is irradiated onto the target to be measured; and adapted to collimate the reflected laser so that the collimated reflected laser enters the single-photon detector; The transmission module comprises: A circulator, adapted to transmit the pulsed laser and to transmit the received collimated reflected laser to the single photon detector; The collimator is used to collimate the pulse laser from the circulator and irradiate the collimated pulse laser onto the target to be measured, and is used to collimate the reflected laser from the target to be measured and transmit it to the circulator.

8. The imaging device according to claim 6, characterized in that The laser module comprises: A signal generator, suitable for sending control signals in a time sequence; A laser is adapted to generate the pulsed laser according to the control signal.

9. The imaging device according to claim 8, characterized in that: The imaging device further comprises: a first optical switch, which allows the pulsed laser to be incident on the target to be measured when in an on state, and prohibits the reflected laser from entering the laser module when in a off state; a second optical switch, which transmits the reflected laser light to the single-photon detector when turned on, and prohibits the pulsed laser light from entering the single-photon detector when turned off; Wherein, in the emission phase of the pulsed laser, the first optical switch is in an open state, and the second optical switch is in a closed state; in the reception phase of the reflected laser, the first optical switch is in a closed state, and the second optical switch is in an open state; The signal generator is also adapted to control the opening and closing of the first optical switch and the opening and closing of the second optical switch.

10. The imaging device according to claim 9, characterized in that The time-to-digital converter is also adapted to determine the emission time of the pulsed laser, and to determine the time when the single-photon detector receives the target photon of the reflected laser according to the electrical signal emitted by the single-photon detector after the pulsed laser is emitted, thereby calculating the flight time of the target photon; The signal generator is also adapted to send out a timing signal while sending out the control signal, so that the time-to-digital converter determines the emission time of the pulse laser according to the timing signal.

Images