Non-vision field imaging method and system

Abstract

The invention provides a non-vision-field imaging method and system, and relates to the field of non-vision-field imaging. The non-vision field imaging method comprises the following steps: transmitting a measurement signal to an intermediate surface; acquiring a plurality of groups of echo signals correspondingly generated on the detection points of the intermediate surface based on the measurement signals in the process of rotating the target to be measured for multiple times; wherein the echo signal is generated by scattering the measurement signal through the intermediate surface, covering the surface of the to-be-measured target, scattering the measurement signal to the intermediate surface by the surface of the to-be-measured target, and scattering the measurement signal again by the intermediate surface; and reconstructing a to-be-detected target image according to the multiple groups of echo signals. By using the non-vision-field imaging method provided by the embodiment of the invention, the non-vision-field to-be-measured target can be efficiently and accurately imaged.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of non-line-of-sight imaging, and in particular, to a non-line-of-sight imaging method and system. BACKGROUND

[0002] Non-line-of-sight imaging technology is a technology for detecting hidden objects behind an obstruction through indirect light signals. Because it can break through the line-of-sight limitation of traditional imaging and achieve non-line-of-sight imaging of hidden objects in a complex obstructed scene, it can be applied to search and rescue, autonomous driving, medical imaging, and other fields.

[0003] However, existing non-line-of-sight imaging methods and technologies are limited by physical models and detection systems, have low resolution, and take a long time to collect data. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a non-line-of-sight imaging method and system. The non-line-of-sight imaging method fully covers a target object by emitting an expanded beam laser, collects one-dimensional depth projection echo information at different angles of the target by performing relative angular displacement, and reconstructs a high-resolution image of the target to be measured by a tomographic imaging algorithm, thereby having the advantages of high imaging resolution and fast imaging speed.

[0005] In a first aspect, the embodiments of the present application provide a non-line-of-sight imaging method, which includes: emitting a measurement signal to an intermediate surface; obtaining a plurality of sets of echo signals corresponding to the measurement signal on a detection point on the intermediate surface in the process of multiple rotations of a target to be measured; wherein the echo signal is generated by the measurement signal scattering on the surface of the target to be measured, being scattered to the intermediate surface by the surface of the target to be measured, and being scattered again by the intermediate surface; and reconstructing an image of the target to be measured according to the plurality of sets of echo signals.

[0006] Optionally, in the embodiments of the present application, the measurement signal includes a signal having a time resolution and capable of scattering on an active surface.

[0007] In the above implementation process, the non-line-of-sight imaging method provided by the embodiments of the present application corresponds the intermediate surface in non-line-of-sight imaging to the expander in reflection tomography, thereby applying the reflection tomographic model to the non-line-of-sight imaging scene. The imaging resolution of the non-line-of-sight imaging method provided by the embodiments of the present application is irrelevant to the detection distance and the aperture of the laser emission / reception optical system, and is only related to the depth resolution of the system. Moreover, the traditional point array scanning is no longer needed, thereby breaking through the resolution and rate limitations of traditional non-line-of-sight imaging.

[0008] Optionally, in the embodiment of the present application, the multiple sets of echo signals corresponding to the detection points on the intermediate surface generated based on the measurement signals in the process of multiple rotations of the target to be measured comprises: controlling the target to be measured to rotate multiple times at a preset angle each time; and obtaining the echo signals generated by the detection points each time until the target to be measured rotates one round.

[0009] Optionally, in the embodiment of the present application, the preset angle ranges from 2° to 10°.

[0010] In the above implementation process, the non-visual imaging method provided by the embodiment of the present application obtains multiple sets of echo signals (obtains data at different projection angles) by controlling the target to be measured to rotate a preset angle each time; the number of projection angles in the non-visual imaging method provided by the embodiment of the present application is significantly lower than the number of pixel arrays in the traditional non-visual imaging method, and the data acquisition efficiency is greatly improved compared with the existing non-visual imaging method.

[0011] Optionally, in the embodiment of the present application, the echo signal is determined according to ; wherein, is the echo signal, is the preset angle, is the distance between the straight line of the detection point along the projection direction and the surface of the target to be measured, and is the reflectivity distribution of the surface of the target to be measured.

[0012] In the above implementation process, the non-visual imaging method provided by the embodiment of the present application uses the reflection tomography technology to take the multi-angle “distance-intensity” echo signals collected by the detector as the line integral projection of the target reflectivity f(x, y), uses the tomographic reconstruction algorithm such as Radon inverse transform to realize high-precision contour reconstruction of the non-visual hidden target to be measured, and the non-visual imaging method provided by the embodiment of the present application can also realize efficient and high-precision image reconstruction of the non-visual target to be measured in a complex environment.

[0013] Optionally, in the embodiment of the present application, the target image is reconstructed according to the multiple sets of echo signals, comprising: based on the multiple sets of echo signals, using a filtered back projection, an iterative algorithm or a deep learning algorithm to reconstruct the target image.

[0014] In the above implementation process, the non-visual imaging method provided by the embodiment of the present application introduces the reflection tomography model into the non-visual imaging, obtains the relative angular displacement between the target to be measured and the detection point by rotating the target to be measured to measure the signal, and then performs image reconstruction of the non-visual target to be measured based on the image reconstruction algorithm; not only can a higher imaging resolution be obtained, but also more efficient non-visual imaging can be realized.

[0015] In a second aspect, the embodiments of the present application provide a non-line-of-sight imaging system, comprising: an intermediate surface, a measurement signal transmitter, a detector, and a processor; the measurement signal transmitter is configured to transmit a measurement signal to the intermediate surface; the detector is configured to obtain a plurality of groups of echo signals generated on a detection point of the intermediate surface based on the measurement signal in a process of multiple rotations of a target object; wherein the echo signal is generated by scattering of the measurement signal on the intermediate surface, covering the surface of the target object, and being scattered to the intermediate surface by the surface of the target object, and then being scattered again by the intermediate surface; and the processor is configured to reconstruct an image of the target object according to the plurality of groups of echo signals.

[0016] Optionally, in the embodiments of the present application, the non-line-of-sight imaging system further comprises a turntable; the turntable is configured to support the target object and drive the target object to rotate.

[0017] Optionally, in the embodiments of the present application, the detector comprises one of a single-photon avalanche diode, an avalanche photodiode, a balanced detector, and a superconducting nanowire single-photon detector.

[0018] Optionally, in the embodiments of the present application, the measurement signal transmitter comprises a laser generator. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0020] Figure 1 a flowchart of the non-line-of-sight imaging method provided by the embodiments of the present application; Figure 2 a schematic diagram of the non-line-of-sight imaging system provided by the embodiments of the present application; Figure 3 a flowchart of the echo signal acquisition provided by the embodiments of the present application; Figure 4 a schematic diagram of the echo signal under a fixed angle provided by the embodiments of the present application; Figure 5 a multi-angle projection signal before image reconstruction provided by the embodiments of the present application; Figure 6 an image schematic diagram after reconstruction of the data in the Figure 5 provided by the embodiments of the present application using a filtered back-projection algorithm; Figure 7 a configuration schematic diagram of an experimental non-line-of-sight imaging system provided by the embodiments of the present application; Figure 8 An example diagram of a first target to be detected provided for an embodiment of the present application; Figure 9 An example diagram of a second target to be detected provided for an embodiment of the present application; Figure 10 An example diagram of a target to be detected provided for an embodiment of the present application; Figure 8 A reconstruction result diagram of a target to be detected in the embodiment of the present application; Figure 11 An example diagram of a target to be detected provided for an embodiment of the present application; Figure 9 A reconstruction result diagram of a target to be detected in the embodiment of the present application. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described below with reference to the drawings of the embodiments of the present application. For example, the flowcharts and block diagrams in the drawings show the possible implementation architecture, functions and operations of the systems, methods and computer program products according to the embodiments of the present application. In this regard, each block in the flowcharts or block diagrams can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders from those noted in the drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and they can also be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and the combination of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, the functional modules in the various embodiments of the present application can be integrated together to form a separate part, or can exist independently, or two or more modules can be integrated to form a separate part.

[0022] Non-visual imaging technology is a technology for detecting hidden objects through indirect light signals. Its characteristics are that the multiple reflected lights of the detection laser emitted by the system after passing through the intermediate surface and the hidden object are combined with a high-precision time-resolved detector and a high-efficiency reconstruction algorithm to realize the imaging of the hidden object.

[0023] Non-line-of-sight (NLOS) imaging can break through the line-of-sight limitation of traditional imaging, and is suitable for complex occluded scenes, and can be applied to search and rescue, autonomous driving, medical imaging, and other fields. Specifically, in autonomous driving, NLOS imaging can detect potential dangers such as pedestrians, vehicles, or road abnormalities in the blind area (such as a curve or behind a large vehicle) that cannot be directly observed by the vehicle, providing early warning and improving the safety and response speed of the autonomous driving system. In search and rescue, NLOS imaging can quickly locate trapped personnel in earthquakes or fires by reflecting signals through intermediate surfaces such as walls. In medical imaging, NLOS imaging can indirectly observe deep organ structures through scattered light on the surface of the skin or tissue, reducing the need for invasive detection, and may be applied to endoscopy or early lesion screening in the future. In industrial manufacturing, NLOS imaging can detect hidden parts in the working environment, enabling defect detection inside closed or complex structures of industrial products.

[0024] NLOS imaging lidar generally obtains information by measuring the time of flight of photons to reconstruct the image of an object hidden from the direct line of sight of the detector. The process is as follows: first, a laser is actively emitted to illuminate the intermediate surface, the laser scatters on the intermediate surface and propagates to the hidden object to undergo a second scattering, and then propagates back to the intermediate surface to undergo a third scattering. The detector receives the photon signal after three scatterings and extracts the photon time of flight information containing the spatial information of the hidden object. Finally, the computer uses this information to reconstruct the image of the hidden object using imaging algorithms.

[0025] In the traditional NLOS imaging method, the laser needs to be scanned on the intermediate surface to obtain more information to better reconstruct the image. This scanning process limits the application scenarios of the technology. First, when performing large-scale point array scanning, a large amount of data acquisition time is consumed, limiting the real-time performance of imaging. Second, scanning relies on a large-area continuous intermediate surface, and when the intermediate surface is irregular or small, it is difficult to obtain enough information to reconstruct the image. Third, the imaging resolution and pixel number are related to the scanning area and point number, and when the number of scanning points is small, the image resolution will also decrease.

[0026] That is, existing NLOS imaging technology is limited by physical models and detection systems, with low resolution and long data acquisition time. Among them, the resolution is mainly limited by the scanning range of the emitted laser on the intermediate surface and the time resolution of the detector, and the data acquisition time is mainly limited by the number of sampling points of the scanning point array.

[0027] Based on this, the application provides a non-visual imaging method and system. The non-visual imaging method uses the reflection tomography laser radar principle to realize non-visual imaging. A one-dimensional depth projection echo information of a target at different angles is collected by emitting an expanded beam laser to completely cover the target object and performing relative angular displacement, and a high-resolution image of the target to be measured is reconstructed by a tomographic imaging algorithm. The non-visual imaging method provided by the embodiments of the application has the advantages of high imaging resolution and fast imaging speed.

[0028] Please refer to Figure 1 and Figure 2 , Figure 1 the flowchart of the non-visual imaging method provided by the embodiments of the application; Figure 2 the schematic diagram of the non-visual imaging system provided by the embodiments of the application.

[0029] In reflection tomography, an expander is used to expand the beam waist diameter of a laser beam to completely cover the target, so as to extract the complete contour information of the object from the echo light signal. The traditional expander generally uses a lens group, but in the non-visual scene of the embodiments of the application (please refer to Figure 2 ), the collimated incident laser is scattered into wide-angle diffuse reflection light after passing through the interface, which can cover the hidden objects, so the interface (for example, a rough wall surface) can replace the function of the expander.

[0030] In the non-visual imaging system provided by the embodiments of the application, the interface of the non-visual scene is used as an expander, so that the physical model of reflection tomography is substituted into the non-visual imaging method provided by the embodiments of the application.

[0031] Please refer to Figure 2 on the basis of Figure 1 , the application provides a non-visual imaging method, which comprises the following steps: Step S100: transmitting a measurement signal to the interface.

[0032] In the above step S100, the measurement signal is first transmitted to the interface. Optionally, the measurement signal includes a signal with time resolution and capable of scattering on the acting surface. For example, pulse laser, laser radar, ultra-wideband acoustic wave and the like can be used as the measurement signal in the embodiments of the application.

[0033] For example, a laser is used to transmit pulse laser to the interface.

[0034] Step S200: obtaining a plurality of groups of echo signals corresponding to the detection points on the interface based on the measurement signal in the process of multiple rotations of the target to be measured.

[0035] Step S300: reconstructing the image of the target object according to the multiple sets of echo signals.

[0036] In the above steps S200 to S300, the multiple sets of echo signals used for reconstructing the image of the target object are obtained. The echo signals are generated by the measurement signal scattering on the surface of the target object and scattering again on the interface after scattering on the surface of the target object.

[0037] In the above implementation process, the measurement signal is diffusely reflected at the irradiation point on the interface, generating wide-angle scattered light propagating into the hidden space, which is equivalent to a collimated laser beam being expanded by the interface into a hemispherical light wave. Part of the scattered light is scattered back after completely covering the hidden target object. Due to the shielding of the shield, the detector cannot directly collect the scattered light returned from the target object, but the detector can be aimed at a detection point on the interface to collect the light returned after the scattered light is scattered again on the interface. The time-of-flight information of the light scattered three times still carries the profile information of the object.

[0038] The echo at a single angle is not enough to reconstruct the image of the object, and therefore, the embodiment of the present application obtains the echo signal obtained after multiple rotations of the target object. For example, the target object is controlled to rotate at least one round to obtain the echo signal at all angles, so as to use the reconstruction algorithm of reflective tomography to reconstruct the profile of the hidden target object.

[0039] In the non-visual imaging method provided by the embodiment of the present application, the lateral resolution is (the same as the theoretical lateral resolution of reflective tomography), while the lateral resolution in the traditional non-visual imaging method is , wherein w is the half width of the scanning area of the interface, z is the distance between the hidden target object and the interface, c is the speed of light, is the time resolution of the detection system.

[0040] It can be seen that , the resolution of the traditional non-visual imaging is limited by w , z and , while the resolution of the non-visual imaging based on reflective tomography provided by the embodiment of the present application is limited by , so the theoretical resolution is higher.

[0041] by Figure 1 and Figure 2It can be known that the non-visual imaging method provided in the embodiment of the present application corresponds the intermediate surface in the non-visual imaging to the expander in the reflection tomography, so that the reflection tomography model is applied to the non-visual imaging scene, the imaging resolution of the non-visual imaging method provided in the embodiment of the present application is irrelevant to the detection distance and the laser emission / reception optical system aperture, and is only related to the depth resolution of the system, and the traditional point array scanning is no longer needed, so that the resolution and rate limit of the traditional non-visual imaging are broken through.

[0042] Please refer to Figure 3 , Figure 3 The echo signal acquisition flowchart is provided in the embodiment of the present application; in the optional embodiment of the embodiment of the present application, the above-mentioned step S200 in which a plurality of groups of echo signals corresponding to the detection points on the intermediate surface based on the measurement signals are generated in the process that the target to be measured rotates multiple times can be realized through the following steps: Step S210: controlling the target to be measured to rotate multiple times at a preset angle each time.

[0043] Step S220: acquiring the echo signal generated by the detection point each time until the target to be measured rotates one round.

[0044] In the above-mentioned step S210 to step S220, the target to be measured is controlled to rotate multiple times at a preset angle each time, wherein for the rotation of the target to be measured, the target object can be placed on a rotating table, and the rotation of the target to be measured is realized by controlling the rotating table.

[0045] After each rotation, the profile of the target to be measured under the laser coverage changes, and by continuously rotating, new profile data is continuously provided.

[0046] In the embodiment of the present application, the selection of the preset angle needs to balance the imaging resolution and the data acquisition efficiency, the smaller the angle is, the denser the projection data is, and the higher the reconstruction accuracy is. For example, if the target needs to rotate 72 times (5° each time) to rotate one round, 72 groups of echo signals need to be collected.

[0047] In an optional embodiment, the preset angle ranges from 2° to 10°.

[0048] Exemplarily, the data acquisition efficiency of the embodiment of the present application is briefly described by taking 5° interval as an example: The non-visual field imaging method based on the reflection tomography principle provided in the embodiments of the present application is related to only the time unit number of single sampling of the pixel points. Currently, a single-angle 256-time unit length signal recording can be conveniently realized by using a single photon avalanche diode to receive a photon signal and using a time-dependent single photon counter to record the signal. As long as the angle information of the target to be measured rotating one round is collected, a 256x256 pixel image can be reconstructed by using an algorithm. Assuming that the sampling interval is 5°, 72 angles are needed for sampling 360°, and the sampling time of each angle is set to 1s. Only 72s is needed to complete data collection.

[0049] However, the traditional non-visual field imaging method uses a point scanning mode, and the pixel points of imaging are consistent with the number of scanning points. Therefore, to obtain a better imaging effect, a large number of points need to be scanned. If a 256x256 dot matrix is scanned, and the scanning time of each point is 1s, a total of about 1092 minutes is needed for scanning. Even if the scanning time of each point is reduced to 0.1s, about 109 minutes is still needed, and such a low scanning time will reduce the signal-to-noise ratio of imaging.

[0050] Therefore, the number of projection angles in the non-visual field imaging method provided in the embodiments of the present application is significantly lower than the number of dot matrix scanning pixels in the traditional non-visual field imaging method. Generally, only 72 angles need to be sampled to have a good reconstruction effect. If the sampling time of each angle is still 1s, data collection can be completed in only 1.2 minutes, and the data collection rate is improved by 910 times compared with the traditional method.

[0051] Exemplarily, when the sampling interval is set to 2°, 180 sampling points are needed to complete 360° full-angle scanning. If the sampling time of each angle is maintained at 1s, the total collection time needs to be 180s (3 minutes). Compared with the 5° interval scheme, the data volume of the 2° interval scheme is increased by 2.5 times, but the angular resolution of the reconstructed image can be significantly improved, and the data collection time is still much lower than that of the traditional non-visual field imaging method.

[0052] Exemplarily, when a 10° sampling interval is used, only 36 sampling points are needed to complete 360° full-angle scanning. Under the condition of maintaining the sampling time of each angle at 1s, the total collection time is shortened to 36s. Compared with the denser sampling scheme, the 10° interval greatly improves the data collection efficiency under the premise of ensuring the basic imaging quality, and is particularly suitable for application scenarios with high real-time requirements, such as dynamic target monitoring or rapid security inspection.

[0053] By Figure 3It can be known that the non-visual imaging method provided in the embodiment of the present application obtains multiple sets of echo signals (acquires data under different projection angles) by controlling the target to be measured to rotate a preset angle each time. The number of projection angles in the non-visual imaging method provided in the embodiment of the present application is significantly lower than the number of point array scanning pixels in the traditional non-visual imaging method, and the data acquisition efficiency is greatly improved compared with the existing non-visual imaging method.

[0054] In an optional embodiment, the echo signal is determined according to .

[0055] Wherein, is the echo signal, is the preset angle, is the distance between the straight line of the detection point along the projection direction and the surface of the target to be measured, is the reflectivity distribution of the surface of the target to be measured.

[0056] In reflectance tomography, the echo signal of the target to be measured measured by the detector is a number of "distance-intensity" response waveforms, which is actually projection data under different detection angles, and is a kind of echo after the reflection cross section of the target to be measured modulates the light signal.

[0057] In the embodiment of the present application, under a certain angle, the three times scattered photon signals reflected by the projection profile of the hidden target to be measured on the horizontal plane are recorded by the detector as p(r,θ), and this waveform is actually the projection of the target surface reflectivity distribution (or target profile image) along the straight line r in the light direction.

[0058] Further, based on the projection algorithm, the can be obtained, so that the image reconstruction of the target to be measured based on the projection data can be realized.

[0059] It can be known that the non-visual imaging method provided in the embodiment of the present application uses the reflectance tomography imaging technology to take the multi-angle "distance-intensity" echo signals collected by the detector as the line integral projection of the target reflectivity f(x,y), and uses the tomographic reconstruction algorithm such as Radon inverse transform to realize the high-precision profile reconstruction of the non-visual hidden target to be measured. The non-visual imaging method provided in the embodiment of the present application can also realize the high-efficiency and high-precision image reconstruction of the non-visual target to be measured in a complex environment.

[0060] Please refer to Figure 4 , Figure 5 and Figure 6 , Figure 4 , which are the echo signal schematic diagrams under fixed angles provided in the embodiment of the present application; Figure 5 , which is the multi-angle projection signal before image reconstruction provided in the embodiment of the present application;Figure 6 The image after the data in the embodiment of the application is reconstructed using a filtered back-projection algorithm is shown in FIG. 3. In an optional implementation of the embodiment of the application, the step S300 of reconstructing the target image according to the multiple sets of echo signals can be implemented in the following manner: using a filtered back-projection, an iterative algorithm or a deep learning algorithm to reconstruct the target image based on the multiple sets of echo signals. Figure 5

[0061] Taking the image reconstruction based on the filtered back-projection algorithm as an example, the image reconstruction process is described as follows: As shown in FIG. 4, the three times scattered photon signals reflected by the projection profile of the hidden target on the horizontal plane at a certain angle are recorded by the detector as Figure 4 This waveform is actually the projection of the target surface reflectance distribution (or the target profile image) along the straight line in the illumination direction. As changes, the target rotates by angles, and r projection data are obtained; according to the Fourier slice theorem, the target profile image is obtained by using the filtered back-projection method as , wherein m and m represent the Fourier transform and the inverse Fourier transform, respectively, is a slope filter in the frequency domain, thereby realizing the non-visual field imaging based on the reflectance tomography, and show the reconstructed images of the rectangular object realized by the two-dimensional distance -angle Figure 5 data composed of m sets of projection data. Figure 6 Optionally, the image reconstruction based on the echo data in the embodiment of the application can also use an iterative algorithm or a deep learning algorithm. Illustratively, first, a deep neural network (such as U-Net or ResNet) is constructed, a large amount of simulated or measured echo signal-target image paired data is used for training, so that the network learns the end-to-end mapping relationship from the multi-angle echo projection data to the target image; in the inference stage, the newly acquired echo signal is input into the trained network, and the network automatically extracts features and outputs the reconstructed target image. r Through

[0062]

[0063] Through Figure 4 to Figure 6 ​​​It can be seen that the non-line-of-sight imaging method provided in the embodiment of the present application introduces the reflection tomography model into the non-line-of-sight imaging, obtains the relative angular displacement between the target to be measured and the detection point by rotating the target to be measured to perform signal measurement, and then reconstructs the image of the non-line-of-sight target to be measured based on the image reconstruction algorithm; not only can higher imaging resolution be obtained, but also more efficient non-line-of-sight imaging can be achieved.

[0064] Please continue to see Figure 2 The present application provides a non-line-of-sight imaging system. The non-line-of-sight imaging system 100 includes: an intermediary surface 110, a measurement signal transmitter 120, a detector 130 and a processor (not shown in the figure). Figure 2 In the figure, A is the target to be measured, H is the occluder, and the target A is blocked by the occluder H and is not in the field of view.

[0065] The measurement signal transmitter 120 is used to transmit a measurement signal to the intermediate surface.

[0066] Detector 130 is used to obtain multiple sets of echo signals generated at detection points on the intermediary surface based on the measurement signal during multiple rotations of target A. The echo signals are generated when the measurement signal is scattered by intermediary surface 110, covers the surface of target A, is scattered by the surface of target A onto intermediary surface 110, and is further scattered by intermediary surface 110.

[0067] The processor is used to reconstruct an image of the target A to be measured according to multiple groups of echo signals.

[0068] Please continue to see Figure 2 Optionally, the non-line-of-sight imaging system further includes a turntable 140. The turntable 140 is used to support the target A to be measured and drive the target A to be measured to rotate.

[0069] Optionally, the detector 130 may be one of a single photon avalanche diode (SPAD), an avalanche photodiode (APD), a balanced photodetector (BPD), and a superconducting nanowire single photon detector (SNSPD).

[0070] Optionally, the measurement signal transmitter 120 includes a laser generator.

[0071] In an optional embodiment, in the process of obtaining multiple rotations of the target A to be measured, multiple groups of echo signals corresponding to the detection points on the intermediate surface are generated based on the measurement signals, including: the turntable 140 drives the target A to be measured to rotate multiple times at a preset angle each time; the detector 130 obtains the echo signal generated by the detection point each time until the target to be measured rotates one circle.

[0072] In an optional embodiment, the preset angle range is [2°, 10°].

[0073] In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to In an optional embodiment, the echo signals are determined according to

[0074] In an optional embodiment, in the process of reconstructing the image of the target object according to the plurality of groups of echo signals, the processor is specifically configured to: based on the plurality of groups of echo signals, reconstruct the image of the target object by using a filtered back-projection, an iterative algorithm or a deep learning algorithm.

[0075] In an optional embodiment, the measurement signal transmitter 120 is capable of emitting a signal with time resolution and capable of scattering on the surface of the target object.

[0076] In order to verify the non-visual imaging effect of the non-visual imaging method provided in the embodiments of the present application, the inventors carried out experimental verification by using the non-visual imaging system described above. Please refer to Figure 7 、 Figure 8 and Figure 9 , Figure 7 for the experimental non-visual imaging system configuration diagram provided in the embodiments of the present application; Figure 8 for the example diagram of the first target object provided in the embodiments of the present application; Figure 9 for the example diagram of the second target object provided in the embodiments of the present application.

[0077] In Figure 7 , for the convenience of display, no shielding object is placed. A picosecond pulsed laser with a working wavelength of 1550 nm, a repetition frequency of 1 MHz, an average power of 100 mW and a pulse width of 300 ps is used as a light source, a single photon avalanche diode (SPAD) with a time jitter of 80 ps is used as a detector, the distance between the two is 1.5 m, and the distance between the target object and the intermediate wall is 0.5 m. After the pulsed laser is emitted, the incident point of the pulsed laser is incident on the intermediate wall, and after being scattered by the wall, it propagates to the surface of the target object placed on the rotating motorized displacement table. After being scattered by the target, it returns to the detection point of the intermediate wall, and after being scattered again, it propagates to the single-pixel SPAD detector and is received. The one-dimensional time-photon number signal waveform generated by the detection is recorded by a time-correlated single-photon counter with a time resolution of 8 ps and transmitted to a processor such as a computer, and the image is reconstructed by a filtered back-projection method.

[0078] In Figure 8In the figure, a vertically placed rectangular object with a length of 14.5 cm, a width of 10.5 cm, and a height of 6.5 cm is used as a hidden target to be tested. Its projection on the horizontal plane is a rectangle of 14.5 cm × 10.5 cm.

[0079] exist Figure 9 In the figure, there is a letter F with a height of 25 cm, a width of 20 cm and a line width of 5 cm placed at 45 degrees to the horizontal plane as a hidden target to be tested. Its projection on the horizontal plane is a proportionally reduced F.

[0080] Based on the above system, we obtained Figure 10 and Figure 11 The reconstruction results shown are Figure 10 Provided in the embodiments of this application Figure 8 Schematic diagram of the reconstruction result of the target to be measured; Figure 11 Provided in the embodiments of this application Figure 9 Schematic diagram of the reconstruction results of the target to be measured; the total time resolution is about 300ps, and when the maximum imaging field of view is 40cm×40cm, the corresponding traditional non-line-of-sight imaging resolution is 12cm, which is larger than the minimum size of the two objects and the imaging will not be distinguishable. The resolution of the reflection tomography non-line-of-sight imaging is 4.5cm, which is smaller than the minimum size of the two objects. Therefore, the imaging of the contour is achieved, which shows that the non-line-of-sight imaging based on reflection tomography has a resolution advantage. During the experiment, each object was driven by an electric translation stage to rotate 360° around an axis perpendicular to the horizontal plane. The signal was sampled every 5°, and the sampling time for each angle was 1s, with a total time of 1.2 minutes. The signal length has 256 time units, so a 256×256 pixel image can be reconstructed. If traditional point scanning imaging is used, and each point is also scanned for 1s, it will take 1092 minutes to image an image with the same number of pixels, which shows that the non-line-of-sight imaging based on reflection tomography provided in the embodiment of the present application has a sampling rate advantage.

[0081] In this document, relational terms such as first and second, etc., are used only 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 "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the elements.

[0082] The above merely provides an example of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A non-line-of-sight imaging method, characterized in that: The non-line-of-sight imaging method comprises: transmitting a measurement signal to the intermediary surface; Acquiring multiple groups of echo signals corresponding to the detection points on the intermediary surface based on the measurement signal during multiple rotations of the target to be measured; wherein the echo signals are generated by the measurement signal being scattered by the intermediary surface, covering the surface of the target to be measured, being scattered by the surface of the target to be measured to the intermediary surface, and being scattered again by the intermediary surface; An image of the target to be measured is reconstructed according to the multiple groups of echo signals.

2. The method according to claim 1, characterized in that The acquiring of a plurality of groups of echo signals correspondingly generated at the detection points of the intermediate surface based on the measurement signal during the multiple rotations of the target to be measured includes: Controlling the target to be measured to rotate multiple times at a preset angle each time; The echo signal generated by the detection point is obtained each time until the target to be measured rotates one circle.

3. The method according to claim 2, characterized in that in, The range of the preset angle is [2°, 10°].

4. The method according to claim 2, characterized in that in, The echo signal is based on Determine; among them, is the echo signal, is the preset angle, The detection point along the projection direction The distance between the straight line and the surface of the target to be measured, is the reflectivity distribution of the surface of the target to be measured.

5. The method according to claim 1, wherein The step of reconstructing an image of the target to be measured according to the plurality of groups of echo signals comprises: Based on the multiple groups of echo signals, the image of the target to be measured is reconstructed using filtered back projection, an iterative algorithm or a deep learning algorithm.

6. The method according to claim 1, characterized in that in, The measurement signal includes a signal having time resolution and capable of being scattered on the active surface.

7. A non-line-of-sight imaging system, characterized in that: The non-line-of-sight imaging system includes: an intermediary surface, a measurement signal transmitter, a detector, and a processor; The measurement signal transmitter is used to transmit a measurement signal to the intermediary surface; The detector is configured to obtain a plurality of echo signals corresponding to the detection points on the intermediary surface based on the measurement signal during multiple rotations of the target to be measured; wherein the echo signals are generated by the measurement signal being scattered by the intermediary surface, covering the surface of the target to be measured, being scattered by the surface of the target to be measured onto the intermediary surface, and being scattered again by the intermediary surface; The processor is used to reconstruct an image of the target to be measured according to the multiple groups of echo signals.

8. The non-line-of-sight imaging system according to claim 7, wherein: The non-line-of-sight imaging system further includes a turntable; The turntable is used to support the target to be measured and drive the target to be measured to rotate.

9. The non-line-of-sight imaging system according to claim 7, wherein: The detector includes one of a single-photon avalanche diode, an avalanche photodiode, a balanced detector, and a superconducting nanowire single-photon detector.

10. The non-line-of-sight imaging system according to claim 7, wherein: The measurement signal transmitter includes a laser generator.

Images