Imaging method and system

Abstract

The embodiment of the present application provides an imaging method and system. The imaging method comprises the following steps: using pseudo-thermal light as probe light, and irradiating the probe light to a moving object to form reflected light; receiving the reflected light by using a single-pixel detector, and determining a Doppler frequency shift distribution of the reflected light relative to the probe light by using a frequency analysis system; obtaining a velocity sequence of the moving object based on the Doppler frequency shift distribution of the reflected light relative to the probe light; and constructing a full-view velocity image and a spatial image of the moving object by using a pre-stored reference spatial thermal light field distribution and the velocity sequence based on a correlation imaging algorithm, wherein the spatial image corresponds to a reflectivity distribution of the moving object, the direction of the velocity of the moving object is perpendicular to the plane where the collection surface of the single-pixel detector is located, and the velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of imaging, and more particularly, to an imaging method and an imaging system. BACKGROUND

[0002] Single-pixel imaging is a technique for obtaining object information using a single-pixel detector. The single-pixel detector has important application prospects in the case of lacking array detectors or being unable to use array detectors. In addition, single-pixel imaging uses a non-point scanning imaging mode, and thus can cover the entire field of view, and has advantages of higher quantum efficiency, lower dark noise, and shorter response time compared with array detectors.

[0003] In addition to obtaining phase, spectral (color), and dimensional information, single-pixel imaging can also obtain velocity information of an object. The acquisition of object velocity information has great significance in the fields of atmospheric wind field measurement, explosion, wind tunnel, transportation, security, biological medicine, and national defense.

[0004] Therefore, there is an urgent need for a method and system with imaging efficiency and imaging quality to obtain, for example, a full-view velocity image and a spatial image of a moving object in a scene. SUMMARY

[0005] Embodiments of the present application provide an imaging method and system that can at least partially solve the above problems in the related art or other problems in the field.

[0006] In one aspect, the present application provides an imaging method, which includes: using a pseudo-thermal light as a probe light, and irradiating the probe light to a moving object to form a reflected light; receiving the reflected light by using a single-pixel detector, and determining a Doppler frequency shift distribution of the reflected light relative to the probe light by using a frequency analysis system; obtaining a velocity sequence of the moving object based on the Doppler frequency shift distribution; and constructing a full-view velocity image and a spatial image of the moving object based on a correlation imaging algorithm, by using a pre-stored reference spatial thermal light field distribution and the velocity sequence, wherein the spatial image corresponds to a reflectivity distribution of the moving object; a direction of the velocity of the moving object is perpendicular to a plane where a collection surface of the single-pixel detector is located; and the velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution.

[0007] In one embodiment of the present application, receiving the reflected light by using a single-pixel detector, and determining a Doppler frequency shift distribution of the reflected light relative to the probe light by using a frequency analysis system includes: performing beam combination of the reflected light and a local oscillator light corresponding to the probe light to form a combined light, collecting the combined light by using the single-pixel detector, and determining the Doppler frequency shift distribution by using the collected combined light.

[0008] In one embodiment of the present application, the reflected light is received by a single-pixel detector, and the Doppler shift distribution of the reflected light relative to the probe light is determined by a frequency analysis system, which includes passing the reflected light through at least one of a grating and a Fabry-Perot interferometer to obtain a frequency distribution of the reflected light, thereby determining the Doppler shift distribution.

[0009] In one embodiment of the present application, the method further includes: sampling the combined light corresponding to each of the pre-stored reference spatial thermal light field distribution multiple times within a collection period to form a sampling sequence; performing Fourier transform on the sampling sequence to form a frequency sequence, and determining the Doppler shift distribution corresponding to each of the pre-stored reference spatial thermal light field distribution based on the frequency sequence to obtain each of the velocity sequence; performing correlation operation between each of the transposed velocity sequence and the pre-stored reference spatial thermal light field distribution to obtain a velocity image slice set corresponding to each of the velocity sequence; and combining the velocity image slice sets to obtain the full field of view velocity image.

[0010] In one embodiment of the present application,

[0011]

[0012] wherein, t is a sampling time point within each collection period; s(t) is the collected combined light; (x, y) is the spatial coordinate of the moving object; I0 is the intensity of the local oscillator light; I s is the intensity of the probe light; P(x, y) is the pre-stored reference spatial thermal light field distribution; T(x, y) is the reflection function of the moving object; Δf is the Doppler shift, wherein Δf = f s -f0, f s is the frequency of the reflected light, f0 is the frequency of the local oscillator light; and is the phase difference between the probe light and the local oscillator light.

[0013] In one embodiment of the present application, the method further includes: determining the direct current component of the frequency sequence after performing Fourier transform on the sampling sequence to form the frequency sequence; and based on the slice corresponding to the direct current component in the velocity image slice set as the spatial image after obtaining the velocity image slice set corresponding to each of the velocity sequence.

[0014] In one embodiment of the present application, before the probe light is irradiated to the moving object, the method further includes: frequency shifting the probe light or the local oscillator light, and determining the moving direction of the moving object based on the frequency shift amount and the Doppler shift distribution.

[0015] In an embodiment of the present application, the correlation imaging algorithm comprises at least one of a compressed sensing recovery algorithm, a standard second-order correlation imaging algorithm, and a deep learning algorithm.

[0016] In an embodiment of the present application, the compressed sensing recovery algorithm comprises, but is not limited to, a basis pursuit algorithm, an orthogonal matching pursuit algorithm, an iterative hard thresholding algorithm, a piecewise orthogonal matching pursuit algorithm, a generalized orthogonal matching pursuit algorithm, and a piecewise weak orthogonal matching pursuit algorithm; the standard second-order correlation imaging algorithm comprises, but is not limited to, a differential algorithm, a differential ghost imaging algorithm, a normalized ghost imaging algorithm, and an instantaneous ghost imaging algorithm; and the deep learning algorithm comprises, but is not limited to, a fully connected neural network, a convolutional neural network, and a recurrent neural network.

[0017] Another aspect of the present application provides an imaging system, which comprises: a detection module configured to use pseudo-thermal light as detection light and irradiate the detection light to a moving object to form reflected light; an acquisition module comprising a single-pixel detector and a frequency analysis system, and configured to receive the reflected light by using the single-pixel detector and determine a Doppler frequency shift distribution of the reflected light relative to the detection light by using the frequency analysis system; an analysis module configured to obtain a velocity sequence of the moving object based on the Doppler frequency shift distribution; and an imaging module configured to construct a full-field velocity image and a spatial image of the moving object based on a correlation imaging algorithm, by using a pre-stored reference spatial thermal light field distribution and the velocity sequence, wherein the spatial image corresponds to a reflectivity distribution of the moving object; a direction of the velocity of the moving object is perpendicular to a plane on which an acquisition surface of the single-pixel detector is located; and the velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution.

[0018] In an embodiment of the present application, the acquisition module is further configured to: combine the reflected light and a local oscillator light corresponding to the detection light to form combined light, acquire the combined light by using the single-pixel detector, and determine the Doppler frequency shift distribution by using the acquired combined light.

[0019] In an embodiment of the present application, the acquisition module is further configured to: pass the reflected light through at least one of a grating and a Fabry-Perot interferometer to obtain a frequency distribution of the reflected light, thereby determining the Doppler frequency shift distribution.

[0020] In an embodiment of the present application, the acquisition module is further configured to: sample the combined light corresponding to each of the pre-stored reference spatial thermal light field distribution multiple times in an acquisition period to form a sampling sequence; and perform Fourier transform on the sampling sequence to form a frequency sequence, determine the Doppler shift distribution corresponding to each of the pre-stored reference spatial thermal light field distribution based on the frequency sequence to obtain each of the velocity sequence; and the imaging module is further configured to: perform correlation operation on each of the transposed velocity sequence and the pre-stored reference spatial thermal light field distribution to obtain a velocity image slice set corresponding to each of the velocity sequence; and combine the velocity image slice sets to obtain the full field of view velocity image.

[0021] In an embodiment of the present application,

[0022]

[0023] wherein t is a sampling time point in each acquisition period; s(t) is the acquired combined light; (x, y) is the spatial coordinate of the moving object; I0is the intensity of the local oscillator light; I s is the intensity of the probe light; P(x, y) is the pre-stored reference spatial thermal light field distribution; T(x, y) is the reflection function of the moving object; Δf is the Doppler shift, wherein Δf = f s -f0, f s is the frequency of the reflected light, f0is the frequency of the local oscillator light; and is the phase difference between the probe light and the local oscillator light.

[0024] In an embodiment of the present application, the acquisition module is further configured to determine the direct current component of the frequency sequence after performing Fourier transform on the sampling sequence to form the frequency sequence; and the imaging module is further configured to, after obtaining the velocity image slice set corresponding to each of the velocity sequence, take the slice corresponding to the direct current component in the velocity image slice set as the spatial image.

[0025] In an embodiment of the present application, the probe module is further configured to perform frequency shift processing on the probe light or the local oscillator light, and determine the motion direction of the moving object based on the frequency shift amount and the Doppler shift distribution.

[0026] In an embodiment of the present application, the correlation imaging algorithm comprises at least one of a compressed sensing recovery algorithm, a standard second-order correlation imaging algorithm, and a deep learning algorithm.

[0027] In one embodiment of the present application, the compressed sensing recovery algorithm includes but is not limited to basis pursuit algorithm, orthogonal matching pursuit algorithm, iterative hard thresholding algorithm, piecewise orthogonal matching pursuit algorithm, generalized orthogonal matching pursuit algorithm, piecewise weak orthogonal matching pursuit algorithm.

[0028] The standard second-order correlation imaging algorithm includes but is not limited to difference algorithm, difference ghost imaging algorithm, normalized ghost imaging algorithm, instantaneous ghost imaging algorithm; and the deep learning algorithm includes but is not limited to fully connected neural network, convolutional neural network, recurrent neural network.

[0029] The present application provides an imaging method and system for simultaneously obtaining a spatial image and a full-field velocity image of a moving object using a single-pixel detector, which has high imaging efficiency and quality, improves the ability of a system including a single-pixel detector to obtain object information, and helps to expand the application field and range of single-pixel imaging. BRIEF DESCRIPTION OF DRAWINGS

[0030] Other features, objects, and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the attached drawings. Among other things, the drawings show:

[0031] Figure 1 is a flowchart of an imaging method according to one embodiment of the present application;

[0032] Figure 2 is a partial schematic diagram of an imaging system according to one embodiment of the present application;

[0033] Figure 3 is a schematic diagram of a moving object according to one embodiment of the present application;

[0034] Figure 4 is a schematic diagram of the positional relationship between a light source emitter, a moving object, and a single-pixel detector according to one embodiment of the present application;

[0035] Figure 5 is a schematic diagram of obtaining a velocity sequence of a moving object according to one embodiment of the present application;

[0036] Figure 6 is a schematic diagram of obtaining a full-field velocity image and a spatial image of a moving object according to one embodiment of the present application;

[0037] Figure 7 is a schematic diagram of a full-field velocity image and a spatial image of a moving object obtained at different moving velocities according to one embodiment of the present application; and

[0038] Figure 8 is a block diagram of an imaging system according to one embodiment of the present application. DETAILED DESCRIPTION

[0039] For a better understanding of the present application, various aspects of the present application will be described in more detail with reference to the accompanying drawings. It is to be noted that the detailed description is only examples of exemplary embodiments of the present application and is not intended to limit the scope of the present application in any way. Throughout the specification, like reference numerals refer to like elements. The expression "and / or" includes any and all combinations of one or more of the associated listed items.

[0040] It should be noted that the expressions first, second, third and the like in this description merely serve to distinguish one feature from another feature, and do not indicate any limitation on the features, in particular do not indicate any chronological order.

[0041] In the drawings, the thicknesses of components, sizes, and shapes are slightly adjusted for the sake of convenience in explanation. The drawings are merely examples and are not strictly drawn to scale. As used in this document, the words "substantially", "approximately", and similar expressions are used as terms of approximation and not as terms of degree, and are intended to account for the inherent deviations in a measuring or calculating value that would be recognized by those of ordinary skill in the art.

[0042] It should also be understood that expressions such as "include", "including", "have", "has", "contain" and / or "containing" and the like, are open-ended expressions that are used to specify the presence of stated features, elements, and / or components, but do not preclude the presence or addition of one or more other features, elements, components, and / or combinations thereof. In addition, when expressions such as "at least one of" appear after a list of two or more items, it is meant that any of the listed items can be present, individually or in combination with one or more of the other listed items. Furthermore, when describing embodiments of the present application, the use of "can" means "one or more embodiments of the present application". Also, the word "exemplary" is intended to mean an example or an illustration.

[0043] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It should also be understood that the words "comprise", "comprising", "comprises", "including", "including", "include" and / or "includes" when used in this specification are taken to specify the presence of stated features, elements and / or components but do not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof. Unless specifically set forth herein, words defining degrees of similarity, such as "about", "approximately" and the like, are used to account for inherent deviations in a measuring or calculating value that would be recognized by those of ordinary skill in the art.

[0044] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. In addition, unless explicitly limited or contrary to the context, the specific steps in the methods described in the present application can not be limited to the order described, but can be performed in any order or in parallel. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0045] In addition, when "connected" or "coupled" is used in the present application, it can mean direct contact or indirect contact between the corresponding components, unless there is an explicit definition or can be derived from the context.

[0046] In addition, those skilled in the art can understand that the numbers shown in the drawings and the following text, such as the number of sampling signals, the number of sampling time points, the number of slices included in the velocity image slice set, etc., are only shown for the purpose of illustration, and the present application is not limited thereto. The specific number can be set according to actual needs.

[0047] Figure 1 is a flowchart of an imaging method 1000 according to an embodiment of the present application. As shown in Figure 1 The present application provides an imaging method 1000, which includes:

[0048] S1, using pseudo-thermal light as probe light, and irradiating the probe light to a moving object to form reflected light.

[0049] S2, receiving the reflected light by using a single-pixel detector, and determining the Doppler shift distribution of the reflected light relative to the probe light by using a frequency analysis system.

[0050] S3, obtaining a velocity sequence of the moving object based on the Doppler shift distribution, wherein the direction of the moving velocity is perpendicular to the plane of the collection surface of the single-pixel detector, and the velocity sequence corresponds to a pre-stored reference spatial thermal light field distribution.

[0051] S4, constructing a full-field velocity image and a spatial image of the moving object by using the pre-stored reference spatial thermal light field distribution and the velocity sequence based on a correlation imaging algorithm, wherein the spatial image corresponds to the reflectivity distribution of the moving object.

[0052] The imaging method 1000 will be described in detail below. Figures 1 to 7

[0053] Single-pixel imaging is a technology for obtaining object information by using a single-pixel detector. The single-pixel detector has important application prospects in the case of lacking an array detector or being unable to use an array detector. In addition, single-pixel imaging uses a non-point scanning imaging mode, and thus can cover the entire field of view, and has advantages of higher quantum efficiency, lower dark noise, and shorter response time compared with an array detector. Single-pixel imaging can obtain phase, spectral (color), and dimensional information, and can also obtain velocity information of an object. The acquisition of object velocity information has great significance in the fields of atmospheric wind field measurement, explosion, wind tunnel, transportation, security and defense, biological medicine, and national defense.

[0054] ​According to the relationship between the moving direction of the object and the normal direction of the single-pixel detector collecting surface, the moving speed of the object can be divided into a transverse moving speed and a longitudinal moving speed. The moving direction of the transverse moving speed is parallel to the normal direction of the single-pixel detector collecting surface, and the moving direction of the longitudinal moving speed is perpendicular to the normal direction of the single-pixel detector collecting surface.

[0055] The probe light irradiated on the moving object forms reflected light containing the spatial information of the moving object. Based on the Doppler effect, the reflected light also contains the speed information of the moving object. Thus, combined with the optical frequency analysis and the single-pixel imaging principle, the spatial image of the moving object can be obtained in the process of obtaining the full-field speed image of the moving object, wherein the spatial image of the moving object corresponds to the reflectivity distribution of the moving object, and reflects the spatial information of the moving object.

[0056] The present application provides an imaging method for simultaneously obtaining the spatial image and the full-field speed image of the moving object using the single-pixel detector. The imaging method has high imaging efficiency and imaging quality, improves the ability of the system including the single-pixel detector to obtain object information, and helps to expand the application field and range of the single-pixel imaging.

[0057] Step S1

[0058] The probe light irradiated on the moving object can be pseudo-thermal light which can simulate the statistical properties of the light field of the true thermal light to the greatest extent. The pseudo-thermal light can be obtained by converting the coherent light into pseudo-thermal light after the laser emitted by the laser passes through a medium (glass, metal, film, etc.) with a random pattern in motion. Alternatively, the above-mentioned glass with a pattern can be replaced by frosted glass. The present application does not limit this.

[0059] Alternatively, the laser passes through a light source modulation unit (optical modulator) to form the probe light irradiated on the moving object, wherein the light source modulation unit includes a random matrix generated by, for example, a computer.

[0060] Figure 2 is a partial schematic diagram of an imaging system 2000 according to an embodiment of the present application.

[0061] As shown in Figure 1 and 2 , in an embodiment of the present application, step S1 uses pseudo-thermal light as the probe light, and the formation of the reflected light after the probe light irradiated on the moving object can include, for example: forming the probe light; and forming the reflected light after the probe light irradiated on the moving object.

[0062] As an alternative, a direct detection method can be employed to obtain the Doppler shift distribution of the reflected light relative to the probe light. In other words, in the subsequent step S2, the frequency distribution of the reflected light can be obtained by passing the reflected light through at least one of the grating and the Fabry-Perot interferometer, so as to determine the Doppler shift distribution of the reflected light relative to the probe light.

[0063] As another alternative, a coherent detection method can be employed to obtain the Doppler shift distribution of the reflected light relative to the probe light by means of the reflected light and the local oscillator light corresponding to the probe light. In other words, in the subsequent step S2, the reflected light and the local oscillator light corresponding to the probe light can be combined to form combined light, the combined light can be collected by the single-pixel detector, and the Doppler shift distribution of the reflected light relative to the probe light can be determined by means of the collected combined light. Compared with the direct detection method, the coherent detection method has higher frequency detection accuracy and higher signal-to-noise ratio.

[0064] Therefore, in the following, the process of obtaining the Doppler shift distribution of the reflected light relative to the probe light will be described in detail by taking the coherent detection method as an example. The process can include the step S1 of forming the probe light and the local oscillator light corresponding to the probe light, and the reflected light formed after the probe light irradiates the moving object, and the step S2 of combining the reflected light and the local oscillator light to form combined light, collecting the combined light by the single-pixel detector, and determining the Doppler shift distribution of the reflected light relative to the probe light by means of the collected combined light. However, it should be understood by those skilled in the art that the method of obtaining the Doppler shift distribution of the reflected light relative to the probe light can be selected according to actual needs, and the present application does not limit this.

[0065] Specifically, as shown in FIG. 1, the Doppler shift distribution of the reflected light relative to the probe light can be obtained by means of the reflected light and the local oscillator light corresponding to the probe light. Figure 2As shown, the generating the probe light can include that the light source emitter 100 can emit a single frequency laser (hereinafter referred to as initial light) of, for example, 532 nm, and the initial light is split into two beams by the optical splitting unit 120 (for example, a beam splitter 1) after passing through the optical folding unit 110, one of which passes through the light source modulation unit 200 and can form the probe light, and the other is the local light corresponding to the probe light. Optionally, the light source modulation unit 200 can be a DMD (Digital Micromirror Device). The DMD can work at a switching speed of 20 KHz, thereby showing different random patterns over time. The pre-stored reference spatial thermal light field distribution can be formed by pre-storing the random patterns. Optionally, the size N of the pattern can be 32x32 (1024), and the number of times of measuring the pattern can be M, wherein N and M are positive integers, and N and M can be the same or different. One of the initial light beams irradiates the DMD to form the probe light, and the pre-stored reference spatial thermal light field distribution of the probe light is the same as the pattern sequence formed by the random pattern on the DMD.

[0066] After the probe light passes through the optical unit 210 (optical unit 1, for example, a long-focus telescope), a relatively clear image can be achieved on a long distance, irradiates the moving object 300, and forms reflected light after being reflected by the moving object 300.

[0067] Figure 3 It is a schematic diagram of the moving object 300 according to an embodiment of the present application.

[0068] Optionally, as shown, Figure 3 The moving object 300 can include a target unit 310 and a movable unit 320, wherein the target unit 310 is fixed on a slider and moves at different speeds on a guide rail with the slider.

[0069] In addition, in an embodiment of the present application, before the probe light irradiates the moving object, the imaging method 1000 can further include a step of performing frequency shift processing on the probe light or the local light, so that the movement direction of the moving object can be determined based on the frequency shift amount and the Doppler frequency shift distribution of the probe light in the subsequent step.

[0070] Specifically, the imaging system 2000 can further include a frequency shift device (not shown). The frequency shift device can be, for example, a combination of two acousto-optic frequency shifters, and a frequency shift amount (f r ) of, for example, 7 MHz can be obtained by the combination of the acousto-optic frequency shifters. In other words, by performing frequency shift on the probe light or the local light, the frequency shift center of the probe light or the local light can be offset by f r , so that when the moving object moves forward and reverses at the same speed, the frequency shifts formed can be f r +fd and f r -f d By frequency shifting the probe light or the local light, the moving direction of the moving object can be determined.

[0071] Step S2

[0072] The step S2 of receiving the reflected light by the single-pixel detector and determining the Doppler shift distribution of the reflected light relative to the probe light by the frequency analysis system can comprise: combining the reflected light and the local light corresponding to the probe light to form combined light, collecting the combined light by the single-pixel detector, and determining the Doppler shift distribution of the reflected light relative to the probe light by the collected combined light.

[0073] Alternatively, as an option, in an embodiment of the present application, the reflected light is passed through at least one of a grating and a Fabry-Perot interferometer, and the frequency distribution of the reflected light is obtained to determine the Doppler shift distribution of the reflected light relative to the probe light.

[0074] Figure 4 Fig. 1 is a schematic diagram of the positional relationship among the light source emitter 100, the moving object 300 and the single-pixel detector 400 according to an embodiment of the present application.

[0075] Referring to Figure 4 The light source emitter 100 and the single-pixel detector 400 are located at the same position as shown by the dashed line in the figure, and the moving object 300 moves along the direction of light propagation. The probe light irradiated to the moving object 300 can form reflected light, which has a Doppler shift Δf relative to the probe light. According to the theory of relativity, the Doppler shift Δf of the reflected light relative to the probe light satisfies:

[0076]

[0077] where C is the speed of light, and f0is the frequency of the probe light irradiated to the moving object.

[0078] In combination with Figure 1 , Figure 4 and formula (1), the reflected light of the moving object 300 is combined with the local light at the light splitting unit 310 (e.g., the light splitting mirror 2) to form combined light. After the combined light is converged by the optical unit 320 (the optical unit 2, e.g., the lens), it enters the single-pixel detector 400. Thus, as shown in formula (2), the combined light detected by the single-pixel detector 400 contains the amplitude information and the speed information of the moving object 300.

[0079]

[0080] where t is a sampling time point in each acquisition period; s(t) is the acquired combined light; (x, y) is the spatial coordinate of the moving object; I0is the intensity of the local light; I s is the intensity of the detected light; P(x, y) is the pre-stored reference spatial thermal light field distribution; T(x, y) is the reflection function of the moving object; Af is the Doppler frequency shift, where Af = f s - f0, f s is the frequency of the reflected light, f0is the frequency of the local light; and is the phase difference between the detected light and the local light.

[0081] In addition, the imaging system 2000 can further include an analog to digital converter (A / D converter, not shown) which can convert the acquired analog signal into a digital signal.

[0082] Optionally, the acquired combined light corresponding to each pattern in the pre-stored reference spatial thermal light field distribution is sampled multiple times in the acquisition period, and a sampling sequence S can be formed. For example, the sampling speed of the analog to digital conversion can be 100 MHz, and the total time for sampling each pattern is 10 us, and the number of sampling points of the S sequence corresponding to each pattern is 1000. In other words, t in formula (2) takes 1-1000 time points, and the time interval is 0.01 us.

[0083] The frequency analysis system according to the embodiments of the present application includes at least one of the above-mentioned grating, the Fabry-Perot interferometer, and the device used to determine the Doppler frequency shift distribution using the acquired combined light.

[0084] Step S3

[0085] Figure 5 is a schematic diagram of obtaining a velocity sequence S(v j , i) of the moving object according to an embodiment of the present application.

[0086] As Figure 5 shown, step S3 obtains the velocity sequence of the moving object based on the Doppler frequency shift distribution of the reflected light relative to the detected light, where the direction of the velocity of the moving object is perpendicular to the plane of the acquisition surface of the single-pixel detector, and the velocity sequence corresponding to the pre-stored reference spatial thermal light field distribution can for example include: performing Fourier transform on the sampling sequence S to form a frequency sequence, determining the Doppler frequency shift distribution corresponding to each of the pre-stored reference spatial thermal light field distribution based on the frequency sequence, to obtain each of the velocity sequence.

[0087] Specifically, the pre-stored reference space thermo-optical field distribution (patterns) can include multiple random patterns. Using formula (2), the probe light of the i-th pattern (pattern(i)) can obtain the i-th sampling sequence S through its corresponding combined beam. i (t). As mentioned above, the i-th sampling sequence S i (t) can consist of, for example, 1000 sampling points. The i-th sampling sequence S i Each frequency in (t) can be processed by, for example, a Fast Fourier Transform (FFT) to obtain 1000 bilateral frequencies, which in turn can yield 500 single-sided frequency sequences S. i (f).

[0088] Furthermore, according to formula (2), the amplitude of the Fourier frequency is proportional to the square root of the reflection function T(x,y) of the moving object, therefore the amplitude of the i-th sampling sequence S i (t) can be squared after FFT transformation to obtain a value proportional to the above reflection function.

[0089] Using formula (1), the one-sided frequency sequence S i (f) can be converted into the i-th velocity sequence S i (v). S i (v) can be an M×500 matrix, where M is the number of times the i-th pattern (i) is measured, and 500 is the number of different velocities in the i-th velocity sequence. Furthermore, to obtain a full-field velocity image of a moving object, it is necessary to... i (v) Transpose to obtain S v (i). like Figure 5 As shown in (d), each curve represents the j-th velocity v. j The corresponding sequence S v (i) Repeat the above steps until the complete velocity sequence of the moving object is obtained.

[0090] Furthermore, the imaging method 1000 may also include a method for forming a spatial image of a moving object. This method includes: in step S3, after performing a Fourier transform on the sampled sequence to form a frequency sequence, determining the DC component of the frequency sequence; and in a subsequent step S4, obtaining a spatial image of the moving object from a set of velocity image slices corresponding to each velocity sequence, where the slices corresponding to the DC component are obtained.

[0091] Step S4

[0092] Figure 6 This is a schematic diagram of obtaining a full-field velocity image and spatial image of a moving object according to one embodiment of this application.

[0093] likeFigure 6 As shown in formula (3) below, step S4 constructs the full-field velocity map and the spatial image of the moving object based on the correlation imaging algorithm, using the pre-stored reference spatial thermal light field distribution and the velocity sequence. The constructing can for example include: performing correlation operation between each of the transposed velocity sequence and the pre-stored reference spatial thermal light field distribution P(x, y) to obtain a velocity image slice set o(v j ; x, y) corresponding to each of the velocity sequence, wherein the slice corresponding to the direct current component in the velocity image slice set o(v j ; x, y) is the spatial image of the moving object; and combining the remaining slices in the velocity image slice set o(v j ; x, y) to obtain the full-field velocity map of the moving object, wherein (x, y) is the spatial coordinate of the moving object.

[0094]

[0095] wherein (x, y) is the spatial coordinate of the moving object; P(x, y) is the pre-stored reference spatial thermal light field distribution; is each of the transposed velocity sequence; o(v j ; x, y) is the velocity image slice set corresponding to each of the velocity sequence.

[0096] Optionally, each of the transposed velocity sequence is multiplied by the pre-stored reference spatial thermal light field distribution P(x, y) to obtain the velocity image slice corresponding to the moving velocity v j . When the i-th velocity sequence has 500 different velocities, the velocity image slice set (Imaging Cube) o(v j ; x, y) can include 500 velocity image slices.

[0097] wherein the velocity image slice corresponding to the direct current component is the spatial image of the object, and the set of the remaining 499 velocity image slices can be combined to obtain the full-field velocity map of the moving object.

[0098] In this embodiment, by combining the optical frequency analysis and the single-pixel imaging principle, the spatial image of the moving object can be obtained in the process of obtaining the full-field velocity map of the moving object, wherein the spatial image of the moving object corresponds to the reflectivity distribution of the moving object and reflects the spatial information of the moving object. The present application first proposes an imaging method for simultaneously obtaining the spatial image and the full-field velocity map of the moving object using a single-pixel detector, which has both imaging efficiency and imaging quality, improves the ability of the system including the single-pixel detector to obtain object information, and helps to expand the application field and range of single-pixel imaging.

[0099] Additionally, in one embodiment of the present application, the correlation imaging algorithm can include at least one of a compressed sensing recovery algorithm, a standard second-order correlation imaging algorithm, and a deep learning algorithm.

[0100] Optionally, the compressed sensing recovery algorithm can include, but is not limited to, a basis pursuit algorithm, an orthogonal matching pursuit algorithm, an iterative hard thresholding algorithm, a sub-orthogonal matching pursuit algorithm, a generalized orthogonal matching pursuit algorithm, and a sub-weak orthogonal matching pursuit algorithm; the standard second-order correlation imaging algorithm can include, but is not limited to, a differential algorithm, a differential ghost imaging algorithm, a normalized ghost imaging algorithm, and an instantaneous ghost imaging algorithm; and the deep learning algorithm can include, but is not limited to, a fully connected neural network, a convolutional neural network, and a recurrent neural network.

[0101] Taking the application of the compressed sensing recovery algorithm in the imaging method 1000 as an example, through the compressed sensing recovery algorithm, a high-quality full-field velocity image of a moving object can be obtained at a low sampling rate, wherein the sampling rate β can be defined as:

[0102]

[0103] wherein N is the size of each pattern in the pre-stored reference spatial thermal light field distribution; M is the number of times each pattern is measured; and β is the sampling rate after the compressed sensing recovery algorithm is adopted.

[0104] Specifically, the pre-stored reference spatial thermal light field distribution P(x, y) and the reflection function T(x, y) of the moving object can be reshaped into one-dimensional vectors P and T. Then, M measurements are performed, and the measurement matrix φ can be composed of M P vectors.

[0105] Thus, the jth velocity v j The corresponding sequence can be represented by formula (5). Hereinafter, for the convenience of description, marked as S(v j ).

[0106] S(v j ) = φT(v j ) (5)

[0107] wherein T(v j ) is a one-dimensional vector of the reflection function corresponding to the jth velocity v j .

[0108] According to the compressed sensing theory of images, there is a sparse vector x(v j in the sparse basis, wherein the sparse vector x(v j corresponds to the jth velocity v jThe corresponding one-dimensional vector T(v) of the reflection function j It satisfies formula (6).

[0109] T(v j )=ψx(v j (6)

[0110] Where ψ is an N×N sparse transformation matrix, and x(v j ) is T(v j The compressed vector in the sparse basis.

[0111] Therefore, in step S4, constructing the full field-of-view velocity image of the moving object through the compressed sensing recovery algorithm can be understood as a convex optimization process of solving formula (7).

[0112]

[0113] Where, |x(v j )| is x(v j The first norm of ); ||S(v j )-φψx(v j )‖ is S(v j )-φψx(v j The second norm of ) is λ, where λ is a nonnegative parameter.

[0114] The compressed vector x(v) is obtained through a convex optimization process. j After that, using the compressed vector x(v) j Formula (6) and equation (7) can be used to obtain the velocity v of the j-th velocity. j The corresponding one-dimensional vector T(v) of the reflection function j This allows us to obtain the full field-of-view velocity image and spatial image of the moving object.

[0115] Optionally, when obtaining the full-field velocity image of a moving object using a compressed sensing restoration algorithm, the sampling rate can be 10% to 30%, such as 25% or 30%. This provides both imaging speed and good imaging quality.

[0116] Figure 7 This is a schematic diagram of a spatial image and a full-field velocity image of a moving object obtained at different motion speeds according to one embodiment of this application.

[0117] like Figure 7 As shown, according to at least one embodiment of the imaging method 1000 provided in this application, spatial images and full-field velocity images of moving objects at different speeds can be obtained. Among them, in Figure 7 The velocity of the moving object in (a) is 31.92 cm / s; Figure 7(b) the moving object has a speed of 125.02 cm / s. When the moving object moves at the above speed, the imaging method 1000 provided by at least one embodiment of the present application can obtain a clear and accurate full-field velocity image. In addition, when the moving object has a speed of 2.66 cm / s to 1327.34 cm / s, the imaging method 1000 can obtain a clear and accurate full-field velocity image, and the velocity resolution of the imaging method 1000 can reach 2.66 cm / s.

[0118] Further, in at least one embodiment of the present application, the analog signal collected can be converted into a digital signal by using an analog-to-digital converter, so as to improve the velocity measurement range and the velocity measurement accuracy of the imaging method 1000. For example, the sampling speed of the analog-to-digital conversion can be 100 MHz, and the total time for sampling each pattern is 10 us, and the sampling points of the S sequence corresponding to each pattern is 1000. If an analog-to-digital converter with a sampling speed of 10 GHz is used to sample for 10 us, the number of sampling points obtained is 100,000, and thus the imaging method 1000 can achieve a velocity measurement accuracy of 0.266 mm / s. If an analog-to-digital converter with a sampling speed of 10 GHz is used to sample for 30 us, the number of sampling points obtained is 300,000, and thus the imaging method 1000 can achieve a velocity measurement accuracy of 0.266 mm / s, and a velocity measurement range of 0.0266 cm / s to 3982.02 cm / s (i.e., 143 km / h).

[0119] Therefore, according to the imaging method 1000 provided by at least one embodiment of the present application, the spatial image and the full-field velocity image of the moving object can be obtained simultaneously by the single-pixel detector, and the imaging efficiency and the imaging quality can be compatible, the ability of the system including the single-pixel detector to obtain object information is improved, and the application field and the application range of the single-pixel imaging are expanded.

[0120] Figure 8 is a block diagram of an imaging system 2000 according to one embodiment of the present application.

[0121] In addition, as Figure 8As shown, another aspect of the present application further provides an imaging system 2000. The imaging system 2000 comprises a detection module 2001, an acquisition module 2002, an analysis module 2003, and an imaging module 2004. The detection module 2001 is configured to use pseudo-thermal light as probe light, and irradiate the probe light to a moving object to form reflected light. The acquisition module 2002 comprises a single-pixel detector and a frequency analysis system, and is configured to receive the reflected light by the single-pixel detector, and determine a Doppler frequency shift distribution of the reflected light relative to the probe light by the frequency analysis system. The analysis module 2003 is configured to obtain a velocity sequence of the moving object based on the Doppler frequency shift distribution of the reflected light relative to the probe light. The imaging module 2001 is configured to construct a full-field velocity image and a spatial image of the moving object based on a correlation imaging algorithm, using a pre-stored reference spatial thermal light field distribution and the velocity sequence. The spatial image corresponds to a reflectivity distribution of the moving object. The direction of the velocity is perpendicular to the plane where the acquisition surface of the single-pixel detector is located. The velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution of the probe light.

[0122] The present application provides an imaging system for simultaneously obtaining a spatial image and a full-field velocity image of a moving object using a single-pixel detector. The system has both high imaging efficiency and high imaging quality, and improves the ability of the system including the single-pixel detector to obtain object information, and helps to expand the application field and range of single-pixel imaging.

[0123] Specifically, the probe light is irradiated on the moving object to form reflected light containing spatial information of the moving object. Based on the Doppler effect, the reflected light also contains velocity information of the moving object. Therefore, by combining the optical frequency analysis and the single-pixel imaging principle, the spatial image of the moving object can be obtained in the process of obtaining the velocity image of the moving object. The spatial image of the moving object corresponds to the reflectivity distribution of the moving object, and reflects the spatial information of the moving object.

[0124] As an option, in an embodiment of the present application, the acquisition module 2002 is further configured to use a direct detection method to pass the reflected light through at least one of a grating and a Fabry-Perot interferometer to obtain a frequency distribution of the reflected light, so as to determine the Doppler frequency shift distribution of the reflected light relative to the probe light.

[0125] As another option, in another embodiment of the present application, the acquisition module 2002 is further configured to use a coherent detection method to combine the reflected light and a local oscillator light corresponding to the probe light to form combined light, acquire the combined light by the single-pixel detector, and determine the Doppler frequency shift distribution of the reflected light relative to the probe light by using the acquired combined light. Compared with the direct detection method, the frequency detection accuracy of the coherent detection method is higher, and the signal-to-noise ratio is also higher.

[0126] In addition, when the coherent detection method is adopted, the detection module 2001 is further configured to form the detection light and the local oscillator light corresponding to the detection light, and the reflected light formed after the detection light irradiates the moving object. In addition, those skilled in the art should understand that the acquisition module can be configured according to actual needs, and the present application does not make any limitation.

[0127] Further, when the coherent detection method is adopted, the acquisition module 2002 is further configured to: in the acquisition period, sample the combined light corresponding to each of the pre-stored reference spatial thermal light field distribution multiple times to form a sampling sequence; and perform Fourier transform on the sampling sequence to form a frequency sequence, and determine the Doppler shift distribution corresponding to each of the pre-stored reference spatial thermal light field distribution based on the frequency sequence, to obtain each of the velocity sequence. The imaging module 2004 is further configured to: correlate each of the transposed velocity sequence with the pre-stored reference spatial thermal light field distribution to obtain a set of velocity image slices corresponding to each of the velocity sequence; and combine the set of velocity image slices to obtain a full field of view velocity image of the moving object.

[0128] Optionally, the acquired combined light s(t) satisfies:

[0129]

[0130] Wherein t is the sampling time point in each acquisition period; s(t) is the acquired combined light; (x, y) is the spatial coordinate of the moving object; I0 is the intensity of the local oscillator light; I s is the intensity of the detection light; P(x, y) is the pre-stored reference spatial thermal light field distribution; T(x, y) is the reflection function of the moving object; Δf is the Doppler shift, wherein Δf = f s -f0, f s is the frequency of the reflected light, f0 is the frequency of the local oscillator light; and is the phase difference between the detection light and the local oscillator light.

[0131] In addition, the acquisition module 2002 is further configured to determine the direct current component of the frequency sequence after performing Fourier transform on the sampling sequence to form the frequency sequence; and the imaging module 2004 is further configured to form the spatial image of the moving object based on the slice corresponding to the direct current component in the set of velocity image slices after obtaining the set of velocity image slices corresponding to each of the velocity sequence.

[0132] The s(t) signal in the above formula does not have a direct current term, but low-frequency disturbances of the system and the environment can cause the s(t) signal to have a low-frequency component, which can be understood to be much lower than the Doppler shift distribution. Thus, the acquisition module 2002 can determine the direct current component of the frequency sequence after performing Fourier transform on the sampling sequence to form the frequency sequence, and the imaging module 2004 can use the direct current component to form the spatial image of the moving object.

[0133] Optionally, in an embodiment of the present application, the imaging system 2000 can further include a frequency shift device (not shown). The frequency shift device can be, for example, a combination of two acousto-optic frequency shifters, and a frequency shift amount of, for example, 7 MHz can be obtained through the combination of the acousto-optic frequency shifters. r In other words, by frequency shifting the probe light or the local oscillator light, the frequency shift center of the probe light or the local oscillator light can be offset by f r Thus, when the moving object moves forward and reverses at the same speed, the formed frequency shifts can be represented as f r + f d and f r - f d By frequency shifting the probe light or the local oscillator light, the moving direction of the moving object can be distinguished.

[0134] In addition, in an embodiment of the present application, the correlation imaging algorithm can include at least one of a compressed sensing recovery algorithm, a standard second-order correlation imaging algorithm, and a deep learning algorithm. Optionally, the compressed sensing recovery algorithm can include, but is not limited to, basis pursuit algorithm, orthogonal matching pursuit algorithm, iterative hard thresholding algorithm, piecewise orthogonal matching pursuit algorithm, generalized orthogonal matching pursuit algorithm, and piecewise weak orthogonal matching pursuit algorithm; the standard second-order correlation imaging algorithm can include, but is not limited to, difference algorithm, difference ghost imaging algorithm, normalized ghost imaging algorithm, and instantaneous ghost imaging algorithm; and the deep learning algorithm can include, but is not limited to, fully connected neural network, convolutional neural network, and recurrent neural network.

[0135] The imaging system provided by at least one embodiment of the present application can combine optical frequency analysis and single-pixel imaging principles to obtain a full-field velocity image of a moving object in a scene, thereby improving the ability of a system including a single-pixel detector to obtain object information and helping to expand the application field and range of single-pixel imaging.

[0136] In addition, the imaging system provided by the present application uses the imaging method provided by at least one embodiment of the present application, and thus the content, structure and related effects involved in the imaging method described above can be fully or partially applicable to the imaging system described herein, and therefore the related or similar content will not be described again.

[0137] The above description is merely exemplary of the application and the principles thereof. It is to be understood that the application is not limited to the specific technical features described above and that the scope of protection of the application is not limited to the technical solutions formed by the specific combinations of the technical features described above, but also covers other technical solutions formed by the arbitrary combinations of the technical features described above or their equivalent features without departing from the technical concept. For example, the technical solutions formed by the mutual replacement of the above features and the technical features disclosed in the application (but not limited to) having similar functions.

Claims

1. An imaging method characterized by, The method comprises: using pseudo-thermal light as probe light, and irradiating the probe light to a moving object to form reflected light; receiving the reflected light by a single-pixel detector, and determining a Doppler shift distribution of the reflected light relative to the probe light by a frequency analysis system; obtaining a velocity sequence of the moving object based on the Doppler shift distribution; and constructing a full-view velocity image and a spatial image of the moving object by a correlation imaging algorithm, using a pre-stored reference spatial thermal light field distribution and the velocity sequence, wherein the spatial image corresponds to a reflectivity distribution of the moving object; a direction of the velocity of the moving object is perpendicular to a plane on which a collection surface of the single-pixel detector is located; and the velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution.

2. The method of claim 1, wherein, The receiving the reflected light by a single-pixel detector, and determining a Doppler shift distribution of the reflected light relative to the probe light by a frequency analysis system comprises: combining the reflected light and a local oscillator light corresponding to the probe light to form combined light, collecting the combined light by the single-pixel detector, and determining the Doppler shift distribution by the collected combined light.

3. The method of claim 1, wherein, The receiving the reflected light by a single-pixel detector, and determining a Doppler shift distribution of the reflected light relative to the probe light by a frequency analysis system comprises: passing the reflected light through at least one of a grating and a Fabry-Perot interferometer to obtain a frequency distribution of the reflected light, thereby determining the Doppler shift distribution.

4. The method of claim 2, wherein, The method further comprises: sampling the combined light corresponding to each of the pre-stored reference spatial thermal light field distributions multiple times within a collection period to form a sampling sequence; performing Fourier transform on the sampling sequence to form a frequency sequence, and determining the Doppler shift distribution corresponding to each of the pre-stored reference spatial thermal light field distributions based on the frequency sequence to obtain each of the velocity sequence; performing correlation operation between each of the transposed velocity sequence and the pre-stored reference spatial thermal light field distribution to obtain a velocity image slice set corresponding to each of the velocity sequence; and combining the velocity image slice set to obtain the full-view velocity image.

5. The method according to claim 4, wherein wherein, t is a sampling time point within each acquisition cycle; is the collected combined light; is a spatial coordinate of the moving object; is an intensity of the local light; is an intensity of the probe light; P is the pre-stored reference spatial thermal light field distribution; T is a reflection function of the moving object; is the Doppler shift, wherein , is a frequency of the reflected light, is a frequency of the local light; and is a phase difference between the probe light and the local light.

6. The method of claim 4, wherein, The method further comprises: determining a direct current component of the frequency sequence after performing Fourier transform on the sampling sequence to form the frequency sequence; and based on a slice corresponding to the direct current component in the velocity image slice set as the spatial image after obtaining the velocity image slice set corresponding to each of the velocity sequence.

7. The method of claim 1, wherein, Before irradiating the probe light to the moving object, the method further comprises: performing frequency shift processing on the probe light or a local oscillator light corresponding to the probe light, and determining a moving direction of the moving object based on a frequency shift amount and the Doppler shift distribution.

8. An imaging system characterized by, The system comprises: a detection module configured to use pseudo-thermal light as probe light, and irradiate the probe light to a moving object to form reflected light; a collection module comprising a single-pixel detector and a frequency analysis system, and configured to receive the reflected light by the single-pixel detector and determine a Doppler shift distribution of the reflected light relative to the probe light by the frequency analysis system; an analysis module configured to obtain a velocity sequence of the moving object based on the Doppler shift distribution; and an imaging module configured to construct a full-view velocity image and a spatial image of the moving object based on a correlated imaging algorithm, by using a pre-stored reference spatial thermal light field distribution and the velocity sequence, wherein the spatial image corresponds to a reflectivity distribution of the moving object, a direction of the velocity of the moving object is perpendicular to a plane where a collection surface of the single-pixel detector is located, and the velocity sequence corresponds to the pre-stored reference spatial thermal light field distribution.

9. The system of claim 8, wherein, The collection module is further configured to: combine the reflected light and a local oscillator light corresponding to the probe light to form a combined light, collect the combined light by the single-pixel detector, and determine the Doppler shift distribution by the collected combined light.

10. The system of claim 8, wherein, The collection module is further configured to: pass the reflected light through at least one of a grating and a Fabry-Perot interferometer to obtain a frequency distribution of the reflected light, thereby determining the Doppler shift distribution.

11. The system of claim 9, wherein The collection module is further configured to: sample the combined light corresponding to each of the pre-stored reference spatial thermal light field distributions multiple times within a collection period to form a sampling sequence; and perform a Fourier transform on the sampling sequence to form a frequency sequence, determine the Doppler shift distribution corresponding to each of the pre-stored reference spatial thermal light field distributions based on the frequency sequence, to obtain each of the velocity sequence; and The imaging module is further configured to: perform a correlation operation between each of the transposed velocity sequence and the pre-stored reference spatial thermal light field distribution to obtain a velocity image slice set corresponding to each of the velocity sequence; and combine the velocity image slice set to obtain the full-view velocity image.

12. The system of claim 11, wherein 13. The system of claim 11, wherein wherein, t is a sampling time point within each acquisition cycle; is the collected combined light; is a spatial coordinate of the moving object; is an intensity of the local light; is an intensity of the probe light; P is the pre-stored reference spatial thermal light field distribution; T is a reflection function of the moving object; is the Doppler shift, wherein , is a frequency of the reflected light, is a frequency of the local light; and is a phase difference between the probe light and the local light. The collection module is further configured to determine a direct current component of the frequency sequence after performing the Fourier transform on the sampling sequence to form the frequency sequence. The imaging module is further configured to determine the spatial image based on a slice corresponding to the direct current component in the velocity image slice set after obtaining the velocity image slice set corresponding to each of the velocity sequence. The probe module is further configured to perform a frequency shift processing on the probe light or a local oscillator light corresponding to the probe light, and determine a moving direction of the moving object based on a frequency shift amount and the Doppler shift distribution. ​ 14. The system of claim 8, wherein, ​

Images