A compression delay imaging system and method based on single-cavity double optical comb endogenous closed-loop coding

Abstract

The application discloses a single-cavity double-optical comb endogenous closed loop coding-based compressed time delay imaging system and method, and belongs to the technical field of time stretching imaging. The technical scheme is that the system comprises an adjustable single-cavity double-optical comb source module, and the adjustable single-cavity double-optical comb source module is sequentially connected with a programmable double-comb composite spectrum module and a closed loop feedback and image reconstruction module in an optical path. The programmable double-comb composite spectrum module and the closed loop feedback and image reconstruction module constitute a closed loop control structure. The application is applied to time stretching imaging.

Description

Technical Field

[0001] This invention relates to the field of time-stretch imaging technology, specifically to a compressed time-delay imaging system and method based on single-cavity dual-optical-comb endogenous closed-loop coding. Background Technology

[0002] Time-stretched imaging is one of the mainstream technologies in the field of ultrafast optical imaging. Its core lies in mapping two-dimensional spatial information into a one-dimensional time-domain waveform. Utilizing a single-pixel high-speed photodetector and an electronic analog-to-digital converter (ADC), equivalent frame rates in the MHz to GHz range can be achieved, fundamentally breaking through the electronic bandwidth bottleneck of traditional CCD / CMOS cameras. With this unique advantage, this technology has been widely used in areas such as biological cell imaging, lidar imaging, soliton motion detection, and ultrafast microscopy.

[0003] However, as application scenarios continue to expand, the massive amounts of data generated by time-stretched imaging place a heavy burden on backend storage, transmission, and processing. To address this, researchers have introduced compressed sensing technology, which utilizes the sparsity of signals to reconstruct the original image with only a small number of projection measurements, significantly reducing the amount of data.

[0004] However, current mainstream encoding generation schemes still suffer from three inherent defects: First, electronic modulation schemes based on high-speed spatial light modulators have complex system structures, extremely high synchronization accuracy requirements, and electronic bandwidth limits the encoding rate, making it difficult to match the frame rate requirements of ultrafast imaging; second, passive optical encoding schemes based on random phase plates have fixed and unadjustable encoding patterns, high inter-coherence, and cannot correct system errors, leading to artifacts and loss of detail in the reconstructed images; third, optical encoding schemes based on dual-comb interferometry, while avoiding the electronic bandwidth bottleneck, suffer from poor common-mode noise suppression and low encoding stability, resulting in unstable single-calibration results. Furthermore, single-comb dual-cavity imaging is currently only used for spectral measurements and has not yet been combined with compressed sensing imaging and closed-loop encoding optimization techniques. In summary, existing compressed time-stretching imaging techniques all suffer from encoding quality defects—either the encoding generation mechanism itself is limited by speed or stability, or it cannot achieve active control of encoding quality and global correction of system errors, making it difficult to meet the quality and long-term stability of compressed reconstructed images for practical applications. Summary of the Invention

[0005] To address the technical problem of encoding quality defects in existing time-stretching imaging systems, this invention proposes a compressed time-delay imaging system and method based on single-cavity dual-comb endogenous closed-loop encoding.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: a compressed time delay imaging system based on single-cavity dual-optical comb endogenous closed-loop coding, including an adjustable single-cavity dual-optical comb source module, the adjustable single-cavity dual-optical comb source module including a fiber ring laser resonator, the fiber ring laser resonator being connected to a control component through a coarse wavelength division multiplexer optical path.

[0007] The optical path of the control component away from the coarse wavelength division multiplexer is connected to a programmable dual-comb composite spectrum module. The programmable dual-comb composite spectrum module includes a multiplexer, and the optical path of the multiplexer is connected to a programmable spectral shaper. The programmable spectral shaper of the programmable dual-comb composite spectrum module is also connected to an optical circulator of the closed-loop feedback and image reconstruction module. The first port of the optical circulator is connected to the output port of the programmable spectral shaper, and the second port of the optical circulator is connected to a spatial dispersion component. The spatial dispersion component projects the purely optical endogenous coded pattern generated by the programmable dual-comb composite spectrum module onto the photosensitive surface of the sample or imaging detector through a movable mirror.

[0008] The data output terminal of the imaging detector is connected to the input terminal of the host computer to provide feedback on the measured illumination pattern of the sample surface.

[0009] The optical path of the third port of the optical circulator is connected to a time stretching component. The time stretching component stretches the signal light output from the third port of the optical circulator in time and then inputs it to the host computer through the photoelectric conversion module. The control output terminal of the host computer is connected to the control input terminal of the programmable spectrum shaper.

[0010] Furthermore, the fiber ring laser resonator includes a pump source, the output end of the pump source is optically connected to the pump input end of a wavelength division multiplexer, and the common end of the wavelength division multiplexer is optically connected in sequence to an erbium-doped fiber, an output coupler, an optical isolator, a carbon nanotube saturable absorber, and a transmission-type dual-passband chirped sampling fiber Bragg grating filter.

[0011] The output of the transmissive dual-passband chirped sampling fiber Bragg grating filter is connected to the signal end optical path of the wavelength division multiplexer, thereby forming a closed fiber ring laser resonator.

[0012] Furthermore, the output end of the output coupler is connected to the optical path of the input end of the coarse wavelength division multiplexer. The output end of the coarse wavelength division multiplexer outputs two optical paths, namely the optical frequency comb of the first wavelength channel and the optical frequency comb of the second wavelength channel. The optical frequency comb of the first wavelength channel enters the first polarization controller, and the optical frequency comb of the second wavelength channel passes through the second polarization controller and the adjustable optical attenuator in sequence and is incident on the piezoelectric ceramic actuator. The optical frequency comb output from the first polarization controller and the optical frequency comb output from the piezoelectric ceramic actuator are combined and input to the programmable dual-comb composite spectrum module.

[0013] The control assembly consists of a first polarization controller, a second polarization controller, an adjustable optical attenuator, and a piezoelectric ceramic actuator.

[0014] Furthermore, the spatial dispersion component includes a cylindrical lens, a grating, and a virtual image phase array connected in sequence by optical paths; the cylindrical lens is used to shape the coded pattern beam into a parallel beam; the grating and the virtual image phase array work together to convert the one-dimensional coded pattern into a two-dimensional spatial scanning array to form a structured illumination pattern.

[0015] Furthermore, a removable mirror is positioned between the virtual image phase array and the sample, and can switch between the two states of being moved in and being moved out;

[0016] In the in-place state, the illumination pattern is directly reflected by the in-place mirror onto the photosensitive surface of the imaging detector;

[0017] When removed, the illumination pattern is unobstructed and illuminates the sample.

[0018] Furthermore, the time stretching component includes an erbium-doped fiber amplifier and a dispersion compensation fiber connected in sequence via optical paths; the input end of the erbium-doped fiber amplifier is connected to the third port of the optical circulator via optical path, the output end of the erbium-doped fiber amplifier is connected to the input end of the dispersion compensation fiber via optical path, and the output end of the dispersion compensation fiber is connected to the input end of the photoelectric conversion module via optical path.

[0019] Furthermore, the photoelectric conversion module includes a balanced photodetector and a digital signal processor connected together; the input end of the balanced photodetector is connected to the output end of the dispersion compensation fiber, the output end of the balanced photodetector is electrically connected to the input end of the digital signal processor, and the output end of the digital signal processor is connected to the input end of the host computer.

[0020] Furthermore, the host computer has a built-in compressed sensing reconstruction algorithm, which uses the signal data transmitted by the photoelectric conversion module to complete the compressed sensing reconstruction and restoration of the sample image.

[0021] A compressed time-delay imaging method based on single-cavity dual-optical-comb endogenous closed-loop coding, using the aforementioned compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding, includes the following steps:

[0022] Step S1: Generate two optical frequency combs with a repetition frequency difference through the adjustable single-cavity dual optical comb source module, and actively adjust the relative time delay of the two optical frequency combs through the control and adjustment components;

[0023] Step S2: In the programmable dual-comb composite spectrum module, the two optical frequency combs are combined, and a pure optical endogenous coding pattern is generated by the interference structure in the spectral domain after the beam is combined. An initial amplitude / phase spectrum mask is applied to the combined dual optical frequency combs to complete the initial control of the coding pattern.

[0024] Step S3: Convert the coded pattern into a two-dimensional spatial scanning array through the spatial dispersion component of the closed-loop feedback and image reconstruction module, move it into the movable reflector, acquire the measured illumination pattern of the sample surface and calculate its inter-coherence; if the inter-coherence does not meet the preset low correlation threshold, generate optimization instructions to iteratively adjust the amplitude / phase spectral mask of the coded pattern, re-acquire and verify until the threshold is met, obtain and lock the current coded pattern parameters and the measured matrix;

[0025] Step S4: Remove the removable mirror and project the optimized coded pattern onto the sample. Collect the signal light returned by the sample and pass it through the time stretching component and photoelectric conversion module in sequence to complete the time stretching and photoelectric conversion. Using the locked measured matrix, the sample image is restored by the compressed sensing reconstruction algorithm.

[0026] The advantages of this invention over the prior art are as follows:

[0027] 1. This invention achieves global coarse adjustment of the coding pattern by changing the relative time delay of the two optical frequency combs through the control component, and achieves fine adjustment of the amplitude / phase of the single comb tooth level by combining the programmable spectrum shaper. It can generate illumination patterns with low inter-coherence between columns and realizes the structure of a single cavity dual optical frequency comb that can be adjusted externally. This avoids the bandwidth bottleneck of high-speed electronic modulation and solves the problem that passive optical coding cannot be designed.

[0028] 2. This invention, based on a movable reflector and imaging detector, constructs a closed-loop calibration self-optimization mechanism, abandoning the traditional approach of relying on theoretical matrices and one-time static calibration. Its advantages are: firstly, it can correct systematic errors; secondly, by leveraging the common-mode noise suppression characteristics of a single-cavity dual-optical frequency comb, the calibration validity period is significantly improved.

[0029] 3. The system of this invention forms a complete coding closed loop. The dual optical frequency comb provides stable coherent endogenous coding, the programmable spectral shaper provides active control degrees of freedom, and the feedback link formed by the imaging detector and the host machine provides experimental calibration and iterative optimization decisions. The three work together to form a compressed time delay imaging platform that can be designed, calibrated, and iteratively optimized. Attached Figure Description

[0030] The present invention will be further described below with reference to the accompanying drawings:

[0031] Figure 1 This is a schematic diagram of the compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding of the present invention.

[0032] Figure 2This is a schematic diagram of the single-cavity dual-optical frequency comb module of the present invention, which achieves external adjustment of relative time delay through adjustment components. Figure 2a is a schematic diagram of the change in relative time delay of the two optical frequency combs without the addition of adjustment components between the closed fiber ring laser resonator and the programmable dual-comb composite spectrum module; Figure 2b is a schematic diagram of the change in relative time delay of the two optical frequency combs after the addition of adjustment components between the closed fiber ring laser resonator and the programmable dual-comb composite spectrum module.

[0033] In the diagram: 1 is the pump source, 2 is the wavelength division multiplexer, 3 is the erbium-doped fiber, 4 is the output coupler, 5 is the optical isolator, 6 is the carbon nanotube saturable absorber, 7 is the transmission-type dual-passband chirped sampling fiber Bragg grating filter, 8 is the coarse wavelength division multiplexer, 9 is the first polarization controller, 10 is the second polarization controller, 11 is the adjustable optical attenuator, 12 is the piezoelectric ceramic actuator, 13 is the combiner, 14 is the programmable spectral shaper, 15 is the optical circulator, 16 is the cylindrical lens, 17 is the grating, 18 is the virtual image phase array, 19 is the sample, 20 is the removable mirror, 21 is the imaging detector, 22 is the erbium-doped fiber amplifier, 23 is the dispersion compensation fiber, 24 is the balanced photodetector, 25 is the digital signal processor, 26 is the computer, 101 is the adjustable single-cavity dual-comb source module, 102 is the programmable dual-comb composite spectrum module, and 103 is the closed-loop feedback and image reconstruction module. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] like Figure 1 As shown, the present invention provides a compressed time-delay imaging system based on single-cavity dual-comb endogenous closed-loop coding, comprising: an adjustable single-cavity dual-comb source module 101, wherein the adjustable single-cavity dual-comb source module 101 is sequentially optically connected to a programmable dual-comb composite spectrum module 102 and a closed-loop feedback and image reconstruction module 103; wherein the programmable dual-comb composite spectrum module 102 and the closed-loop feedback and image reconstruction module 103 constitute a closed-loop control structure.

[0037] 1. Adjustable Single-Cavity Dual-Light Comb Source Module 101

[0038] (I) Structural Description

[0039] The adjustable single-cavity dual optical comb source module 101 includes a fiber ring laser resonator, which is used to generate two optical frequency combs (dual optical frequency combs) with a repetition frequency difference. The fiber ring laser resonator is connected to a control component through a coarse wavelength division multiplexer 8. The relative time delay of the two optical frequency combs can be actively controlled by controlling the control component.

[0040] The fiber ring laser resonator includes a pump source 1. The output end of the pump source 1 is connected to the pump input end of a wavelength division multiplexer 2 via fiber fusion splicing. The common end of the wavelength division multiplexer 2 is sequentially connected to an erbium-doped fiber 3, an output coupler 4, an optical isolator 5, a carbon nanotube saturable absorber 6, and a transmission-type dual-passband chirped sampling fiber Bragg grating filter 7. The output end of the transmission-type dual-passband chirped sampling fiber Bragg grating filter 7 is connected to the signal end of the wavelength division multiplexer 2, thereby forming a closed fiber ring laser resonator.

[0041] In this embodiment, the pump source 1 is a 980nm semiconductor laser tube; the erbium-doped fiber 3, as the gain medium, generates 1550nm laser light under the action of the pump source 1; the output coupler 4 is used to extract part of the mode-locked pulses from the ring laser resonator as two optical frequency combs for output; the optical isolator 5 is used to ensure unidirectional optical transmission within the ring laser resonator; the carbon nanotube saturable absorber 6 is used to achieve passive mode-locking; and the transmission-type dual-passband chirped sampling fiber Bragg grating filter 7 is used to select two transmission passbands within the ring laser resonator to achieve two optical frequency combs with reusable wavelengths for output.

[0042] The output optical path of the output coupler 4 is connected to a coarse wavelength division multiplexer 8. The output coupler 4 introduces two optical frequency combs into the coarse wavelength division multiplexer 8, which splits the two optical frequency combs into two separate optical frequency combs: one for the first wavelength channel and the other for the second wavelength channel. The optical frequency comb for the first wavelength channel enters the first polarization controller 9; the optical frequency comb for the second wavelength channel passes sequentially through the second polarization controller 10 and the adjustable optical attenuator 11 before being incident on the piezoelectric ceramic actuator 12. The optical frequency combs output from the first polarization controller 9 and the piezoelectric ceramic actuator 12 are combined and then input to the programmable dual-comb composite spectrum module 102.

[0043] The first polarization controller 9 and the second polarization controller 10 are used to adjust the polarization state of the two optical frequency combs so that the vibration directions of the two optical frequency combs are the same in the same propagation direction, so as to meet the conditions for coherent interference. The adjustable optical attenuator 11 is used to adjust the power of the optical frequency combs and balance the intensity of the two optical frequency combs to improve the contrast of the subsequent coding pattern. The piezoelectric ceramic actuator 12 is used to introduce a controllable additional optical path difference to the optical frequency comb of the second wavelength channel. By periodically adjusting the voltage applied to the piezoelectric ceramic actuator 12, the relative time delay of the two optical frequency combs can be actively controlled. By adjusting the relative time delay when the two optical frequency combs are recombined, the subsequent interference coding pattern can be changed, so as to realize the programmable adjustment of the coding pattern.

[0044] The first polarization controller 9, the second polarization controller 10, the adjustable optical attenuator 11, and the piezoelectric ceramic actuator 12 constitute the control assembly.

[0045] (II) Explanation of Principles

[0046] like Figure 2 As shown, the relative time delay of the two optical frequency combs can be adjusted by applying a voltage to the piezoelectric ceramic actuator 12 of the adjustable single-cavity dual optical comb source module 101. In the figure, This represents the difference in repetition frequency. For the natural time-shift step size, This is a variable time difference. For example... Figure 2 As shown in Figure a, the repetition frequency difference within the two optical frequency combs determines the inherent relative time delay. This amount of slip is usually fixed and cannot meet the needs of subsequent encoding. Figure 2 Section b illustrates the relative time delay of the two optical frequency combs after a control component is added between the closed fiber ring laser resonator and the programmable dual-comb composite spectrum module 102. The function of this external control component is not to alter the original natural relative slip, but rather to construct an artificially controllable additional time delay term. With this additional term, the total effective relative delay can be expressed as: At this point, the total effective relative time delay is no longer determined solely by the natural parameter—the inherent relative time delay—but by the combined contribution of the natural term and the artificial term—the additional time delay term. Consequently, the composite spectrum after beam combining changes from being dominated by fixed parameters to being a result that can be actively adjusted externally. In summary, with the aid of an external control component, the two optical frequency combs can be reconstructed into an adjustable interferometric spectral encoder, thus providing the necessary degrees of freedom for the closed-loop optimization of the subsequent coding pattern.

[0047] II. Programmable Dual-Comb Composite Spectrum Module 102

[0048] (I) Structural Description

[0049] The programmable dual-comb composite spectrum module 102 includes a combiner 13, which is optically connected to a programmable spectral shaper 14. The input of the combiner 13 is optically connected to the output of the adjustable single-cavity dual-comb source module 101, and is used to combine the two optical frequency combs output by the adjustable single-cavity dual-comb source module 101. Due to the slight difference in the repetition frequency of the two combined optical frequency combs, an interference structure is naturally formed in the spectral domain. This interference structure serves as a purely optical endogenous coding pattern. The programmable spectral shaper 14 receives the coding pattern and performs spectral control on its inter-coherence, actively generating a structured illumination code with lower correlation and better suitability for compressed sensing reconstruction.

[0050] (II) Explanation of Principles

[0051] The wave equations for the two optical frequency combs are as follows:

[0052] ;

[0053] ;

[0054] in, and These represent the amplitudes of the two coherent light beams; and These are the repetition frequencies of the two coherent beams, respectively. For time; and These are the electric fields of the two optical frequency combs, respectively.

[0055] After the two optical frequency combs are combined by the combiner 13, the optical field of the two combined optical frequency combs is calculated according to the trigonometric function formula. It can be represented as:

[0056] ;

[0057] The amplitude of the combined optical beam can be derived as follows:

[0058] ;

[0059] Therefore, it can be seen that the amplitude of the optical field of the two combined optical frequency combs exhibits a periodic change, and the frequency of this change is the difference between the repetition frequencies of the two optical frequency combs. This periodic variation in intensity constitutes the purely optical endogenous coding pattern. Because the two optical frequency combs have a comb-like spectral structure, the two combined optical frequency combs form dense interference fringes in the spectral domain, with each comb tooth corresponding to a coding unit, thus forming a high-dimensional coding pattern.

[0060] The mechanism of the programmable spectral shaper 14 is not to replace the intrinsic coding generated by the dual optical frequency comb itself, but to apply a manually set spectral transfer function to it. Specifically, the composite spectrum after combining the two optical frequency combs can be expressed as follows: This already includes all the endogenous encoded information such as the amplitude, phase, relative time delay, and relative power of each comb tooth resulting from the coherent superposition of the two optical frequency combs. The programmable spectral shaper 14 performs secondary modulation of this composite spectrum in the subsequent stage; its function can be expressed as:

[0061] ;

[0062] in, It is the first The spectral tuning function set by the programmable spectral shaper 14 during the secondary encoding process can include amplitude attenuation, phase delay, or specific spectral masks. In other words, the dual optical frequency comb first provides a stable, coherent, and tunable basic composite spectrum, and then the programmable spectral shaper 14 applies designable weights to different wavelength channels, thereby further upgrading the originally naturally generated physical endogenous encoding by the dual optical frequency comb into a designable and optimizable spectral encoding. It should be noted that the spectral tuning function... The specific form is not arbitrarily set, but is automatically generated by the host computer 26 based on the quality evaluation results of the measured coded pattern of the imaging detector 21 in the closed-loop process described later. This closed-loop control mechanism will be explained in detail in the principle section of the closed-loop feedback and image reconstruction module 103.

[0063] III. Closed-Loop Feedback and Image Reconstruction Module 103

[0064] (I) Structural Description

[0065] The closed-loop feedback and image reconstruction module 103 includes an optical circulator 15. The first port of the optical circulator 15 is optically connected to the output of the programmable spectral shaper 14. The second port of the optical circulator 15 is optically connected to a spatial dispersion component. The spatial dispersion component is used to convert the spectral encoding information of the one-dimensional encoded pattern into a two-dimensional spatial scanning array and project it onto the photosensitive surface of the sample 19 or the imaging detector 21 through the movable reflector 20.

[0066] The spatial dispersion component includes a cylindrical lens 16, a grating 17, and a virtual image phase array 18 connected in sequence by optical paths. The cylindrical lens 16 is used to shape the coded pattern beam into a parallel beam. The grating 17 and the virtual image phase array 18 work together to convert the one-dimensional coded pattern into a two-dimensional spatial scanning array to form a structured illumination pattern.

[0067] The removable mirror 20 is positioned between the virtual image phase array 18 and the sample 19, and can switch between two states: removable and removable. In the removable state (encoding optimization stage), the illumination pattern is directly reflected by the removable mirror 20 to the photosensitive surface of the imaging detector 21; in the removable state (sample 19 imaging stage), the illumination pattern is unobstructedly illuminated onto the sample 19.

[0068] The data output terminal of the imaging detector 21 is connected to the input terminal of the host computer 26, which is used to provide feedback on the measured illumination pattern of the sample 19 surface.

[0069] The optical path of the third port of the optical circulator 15 is connected to a time stretching component. The time stretching component stretches the signal light output from the third port of the optical circulator 15 in time and then inputs it to the host computer 26 through the photoelectric conversion module. The control output terminal of the host computer 26 is connected to the control input terminal of the programmable spectral shaper 14. It is used to generate optimization instructions based on the inter-coherence of the measured illumination pattern to control the programmable spectral shaper 14 to iteratively adjust the amplitude / phase spectral mask of the encoded pattern until the encoded pattern meets the preset low correlation threshold. The current measured matrix and spectral mask parameters are locked, and the image of sample 19 is recovered using the locked measured matrix and compressed sensing algorithm.

[0070] The time stretching assembly includes an erbium-doped fiber amplifier 22 and a dispersion-compensating fiber 23 connected in sequence via optical paths. The input end of the erbium-doped fiber amplifier 22 is connected to the third port of the optical circulator 15 via optical path, the output end of the erbium-doped fiber amplifier 22 is connected to the input end of the dispersion-compensating fiber 23 via optical path, and the output end of the dispersion-compensating fiber 23 is connected to the input end of the photoelectric conversion module via optical path.

[0071] The photoelectric conversion module includes a balanced photodetector 24 and a digital signal processor 25 connected together; the input end of the balanced photodetector 24 is connected to the output end of the dispersion compensation fiber 23, the output end of the balanced photodetector 24 is electrically connected to the input end of the digital signal processor 25, and the output end of the digital signal processor 25 is connected to the input end of the host computer 26.

[0072] When the removable mirror 20 is in the removed state, the signal light reflected or scattered by the sample 19 returns along the original illumination path, passes sequentially through the virtual image phase array 18, the grating 17, and the cylindrical lens 16, and then enters from the second port of the optical circulator 15 and exits from the third port of the optical circulator 15. The signal light output from the third port is amplified sequentially by the erbium-doped fiber amplifier 22, time-stretched by the dispersion compensation fiber 23, and then photoelectrically converted by the balanced photodetector 24. The converted electrical signal is acquired and processed by the digital signal processor 25, and the image of the sample 19 is reconstructed in the host computer 26 using the built-in compressed sensing reconstruction algorithm.

[0073] In this embodiment, the imaging detector 21 is a CCD camera.

[0074] (II) Explanation of Principles

[0075] 1. Coding optimization stage

[0076] Before formal imaging, the system first enters the encoding optimization stage, whereby the movable mirror 20 is moved into the optical path, and the illumination pattern is reflected to the imaging detector 21 before being modulated by the sample 19. The host computer 26 uses this pattern to construct the current actual measurement matrix. And calculate the inter-coherence of its columns. As an indicator for evaluating coding quality, the host computer 26 will calculate the... With a pre-set threshold stored in memory (For example, 0.5) for comparison. If If the current encoded pattern is deemed to be of acceptable quality, the system locks the settings of the current programmable spectral shaper 14, exits the encoding optimization mode, and... As the final reconstruction matrix, it is ready to proceed with the normal sample 19 imaging process; if If the correlation of the current encoded pattern is too high, it is deemed unfavorable for reconstruction. In this case, the host computer 26 employs a pre-defined iterative optimization algorithm (e.g., a gradient-based optimization algorithm), whose objective function is... The iterative optimization algorithm is based on the current The values ​​and their gradients are used to calculate a new set of values ​​aimed at reducing... The amplitude / phase spectral mask parameters, i.e., a new spectral transfer function. The host computer 26 will characterize this New electronic control commands (such as a set of voltage values) are sent to the drive circuit of the programmable spectrum shaper 14, causing corresponding changes in its internal optical modulation units (such as liquid crystal pixels). This physically updates the amplitude / phase mask applied to the dual-frequency comb composite spectrum, generating a new optically encoded pattern. This new pattern is also reflected to the imaging detector 21 via the removable mirror 20. The system automatically repeats the above-described acquisition, evaluation, and optimization process until... The conditions are met. Through the above structure and process, the system of the present invention realizes active and automatic optimization of the quality of the coded pattern, ensuring that a low-correlation illumination pattern can be obtained before each imaging, thereby fundamentally guaranteeing the stability and image quality of compressed sensing reconstruction.

[0077] 2. Imaging stage of sample 19

[0078] After the encoding optimization stage is completed, the removable reflector 20 is moved out of the optical path, allowing the optimized illumination pattern to directly illuminate the sample 19. The signal light returning from the sample 19 carries its spatial information, refracts back along the original illumination optical path to the optical circulator 15, and is output from the third port. After being amplified by the erbium-doped fiber amplifier 22, the signal light enters the dispersion compensation fiber 23. Utilizing the dispersion characteristics of the fiber, light signals of different wavelengths are broadened in the time domain, thereby realizing the mapping of two-dimensional spatial information to a one-dimensional time-domain waveform.

[0079] The time-stretched optical signal is converted into an electrical signal by a balanced photodetector 24. This balanced photodetector 24 employs two matched photodiodes and a low-noise transimpedance amplifier, effectively eliminating common-mode noise and improving the signal-to-noise ratio. The converted electrical signal is acquired by a digital signal processor 25 and transmitted to a host computer 26, where the measured matrix obtained and locked during the aforementioned encoding optimization stage is used. The image of sample 19 was recovered using a compressed sensing reconstruction algorithm.

[0080] 3. Mathematical principles of the reconstruction algorithm

[0081] From a mathematical perspective, the compressed sensing model can be represented as: ,in It is a measurement matrix composed of lighting patterns. It is a sparse vector of the original image to be reconstructed; To account for error terms caused by system noise, mechanical noise, and other non-ideal factors; This is the compressed measurement value obtained from the balanced photodetector 24. In traditional compressed sensing systems where there is no imaging detector 21 for calibration, an ideal measurement matrix is ​​typically used. To reconstruct the model, but in actual optical systems, various non-ideal factors such as spatial dispersion error, aberrations, spectral shaping bias, power drift, and optical path instability can lead to changes in the actual measurement matrix during the measurement process. Deviation from the ideal value, i.e., the ideal measurement matrix With the actual measurement matrix A certain difference will appear between them. If an ideal measurement matrix is ​​still used at this point... If image reconstruction is performed using this difference, it will directly introduce reconstruction error, manifesting as reconstruction artifacts, blurred edges, and loss of detail. These negative effects are particularly pronounced at high compression ratios. The system of this invention utilizes an imaging detector 21 to perform calibration. The imaging detector 21 directly captures the illumination pattern located at the conjugate position of the sample 19 surface. Since this pattern is not modulated by the sample 19, the measured matrix... It accurately reflects the distribution of the actual lighting light field and satisfies... Therefore, the matrix measured using imaging detector 21 As the measurement matrix used for reconstruction, the reconstruction model then becomes In this way, the measurement matrix used by the reconstruction algorithm is completely consistent with the actual lighting process, thereby eliminating the errors previously caused by matrix bias and significantly improving the stability and reliability of the compressed reconstruction results.

[0082] A compressed time-delay imaging method based on single-cavity dual-optical-comb endogenous closed-loop coding, using the aforementioned compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding, includes the following steps:

[0083] Step S1: Two optical frequency combs with a repetition frequency difference are generated by the adjustable single-cavity dual optical comb source module 101, and the relative time delay of the two optical frequency combs is actively adjusted by the control and regulation component.

[0084] Step S2: In the programmable dual-comb composite spectrum module 102, the two optical frequency combs are combined, and a pure optical endogenous coding pattern is generated by the interference structure in the spectral domain after the beam is combined. An initial amplitude / phase spectrum mask is applied to the combined dual optical frequency combs to complete the initial control of the coding pattern.

[0085] Step S3: The coded pattern is converted into a two-dimensional spatial scanning array through the spatial dispersion component of the closed-loop feedback and image reconstruction module 103. The movable reflector 20 is moved in to acquire the measured illumination pattern of the sample 19 and calculate its inter-coherence. If the inter-coherence does not meet the preset low correlation threshold, an optimization instruction is generated to iteratively adjust the amplitude / phase spectrum mask of the coded pattern, and the acquisition and verification are repeated until the threshold is met. The current coded pattern parameters and the measured matrix are obtained and locked.

[0086] Step S4: Remove the removable reflector 20 and project the optimized coded pattern onto the sample 19. Collect the signal light returned by the sample 19 and pass it through the time stretching component and photoelectric conversion module in sequence to complete the time stretching and photoelectric conversion. Using the locked measured matrix, the image of the sample 19 is recovered by the compressed sensing reconstruction algorithm.

[0087] Regarding the specific structure of this invention, it should be noted that the connection relationships between the various component modules used in this invention are definite and achievable. Except as specifically described in the embodiments, their specific connection relationships can bring about corresponding technical effects and solve the technical problems proposed by this invention without relying on the execution of corresponding software programs. The models of the components, modules, and specific components appearing in this invention, the connection methods between them, and the conventional usage methods and expected technical effects brought about by the above technical features, unless specifically described, are all publicly disclosed content in patents, journal articles, technical manuals, technical dictionaries, and textbooks that can be obtained by those skilled in the art before the application date, or belong to conventional technology, common knowledge, and other existing technologies in this field. There is no need to elaborate, which makes the technical solution provided in this case clear, complete, and achievable, and can reproduce or obtain corresponding physical products based on this technical means.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding, characterized in that, The tunable single-cavity dual-optical-comb source module (101) includes a fiber ring laser resonator, which is connected to a control component via a coarse wavelength division multiplexer (8). The optical path of the control component away from the coarse wavelength division multiplexer (8) is connected to a programmable dual-comb composite spectrum module (102). The programmable dual-comb composite spectrum module (102) includes a combiner (13). The optical path of the combiner (13) is connected to a programmable spectral shaper (14). The programmable spectral shaper (14) of the programmable dual-comb composite spectrum module (102) is also connected to an optical circulator (15) of the closed-loop feedback and image reconstruction module (103). The first port of the optical circulator (15) is connected to the output port of the programmable spectral shaper (14). The second port of the optical circulator (15) is connected to a spatial dispersion component. The spatial dispersion component projects the pure optical endogenous coding pattern generated by the programmable dual-comb composite spectrum module (102) onto the photosensitive surface of the sample (19) or the imaging detector (21) through a movable mirror (20). The data output terminal of the imaging detector (21) is connected to the input terminal of the host computer (26) for feedback of the measured illumination pattern on the sample (19) surface; The optical path of the third port of the optical circulator (15) is connected to a time stretching component. The time stretching component stretches the signal light output from the third port of the optical circulator (15) in time and then inputs it to the host computer (26) through the photoelectric conversion module. The control output terminal of the host computer (26) is connected to the control input terminal of the programmable spectrum shaper (14).

2. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 1, characterized in that, The fiber ring laser resonator includes a pump source (1), the output end of the pump source (1) is connected to the pump input end of a wavelength division multiplexer (2), and the common end of the wavelength division multiplexer (2) is connected in sequence to an erbium-doped fiber (3), an output coupler (4), an optical isolator (5), a carbon nanotube saturable absorber (6), and a transmission-type dual-passband chirped sampling fiber Bragg grating filter (7). The output end of the transmission-type dual-passband chirped sampling fiber Bragg grating filter (7) is connected to the signal end optical path of the wavelength division multiplexer (2) to form a closed fiber ring laser resonator.

3. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 2, characterized in that, The output end of the output coupler (4) is connected to the optical path of the input end of the coarse wavelength division multiplexer (8). The output end of the coarse wavelength division multiplexer (8) outputs two optical paths, namely the optical frequency comb of the first wavelength channel and the optical frequency comb of the second wavelength channel. The optical frequency comb of the first wavelength channel enters the first polarization controller (9), and the optical frequency comb of the second wavelength channel passes through the second polarization controller (10) and the adjustable optical attenuator (11) in sequence and is incident on the piezoelectric ceramic actuator (12). The optical frequency comb output from the first polarization controller (9) and the optical frequency comb output from the piezoelectric ceramic actuator (12) are combined and input to the programmable dual-comb composite spectrum module (102). The first polarization controller (9), the second polarization controller (10), the adjustable optical attenuator (11), and the piezoelectric ceramic actuator (12) constitute the control assembly.

4. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 1, characterized in that, The spatial dispersion component includes a cylindrical lens (16), a grating (17), and a virtual image phase array (18) connected in sequence by optical paths; the cylindrical lens (16) is used to shape the coded pattern beam into a parallel beam; the grating (17) and the virtual image phase array (18) work together to convert the one-dimensional coded pattern into a two-dimensional spatial scanning array to form a structured illumination pattern.

5. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 4, characterized in that, A movable mirror (20) is positioned between the virtual image phase array (18) and the sample (19), and can switch between the two states of being moved in and being moved out; In the in-place state, the illumination pattern is directly reflected by the in-place mirror (20) to the photosensitive surface of the imaging detector (21); In the removed state, the illumination pattern is illuminated onto the sample without obstruction (19).

6. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 1, characterized in that, The time stretching component includes an erbium-doped fiber amplifier (22) and a dispersion compensation fiber (23) connected in sequence by optical paths; the input end of the erbium-doped fiber amplifier (22) is connected to the third port of the optical circulator (15) by optical path, the output end of the erbium-doped fiber amplifier (22) is connected to the input end of the dispersion compensation fiber (23) by optical path, and the output end of the dispersion compensation fiber (23) is connected to the input end of the photoelectric conversion module by optical path.

7. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 6, characterized in that, The photoelectric conversion module includes a balanced photodetector (24) and a digital signal processor (25) connected together; the input end of the balanced photodetector (24) is connected to the output end of the dispersion compensation fiber (23), the output end of the balanced photodetector (24) is electrically connected to the input end of the digital signal processor (25), and the output end of the digital signal processor (25) is connected to the input end of the host computer (26).

8. The compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding according to claim 1, characterized in that, The host computer (26) has a built-in compressed sensing reconstruction algorithm. The compressed sensing reconstruction algorithm uses the signal data transmitted by the photoelectric conversion module to complete the compressed sensing reconstruction and restoration of the sample (19) image.

9. A compressed time-delay imaging method based on single-cavity dual-optical-comb endogenous closed-loop coding, characterized in that, Using a compressed time-delay imaging system based on single-cavity dual-optical-comb endogenous closed-loop coding as described in any one of claims 1-8, the method includes the following steps: Step S1: Generate two optical frequency combs with a repetition frequency difference through the adjustable single-cavity dual optical comb source module (101), and actively adjust the relative time delay of the two optical frequency combs through the control and adjustment components; Step S2: In the programmable dual-comb composite spectrum module (102), the two optical frequency combs are combined, and a pure optical endogenous coding pattern is generated by the interference structure of the spectral domain after the combination. An initial amplitude / phase spectrum mask is applied to the combined dual optical frequency comb to complete the initial control of the coding pattern. Step S3: Convert the coded pattern into a two-dimensional spatial scanning array through the spatial dispersion component of the closed-loop feedback and image reconstruction module (103), move it into the movable reflector (20), collect the measured illumination pattern of the sample (19) surface and calculate its inter-coherence; if the inter-coherence does not meet the preset low correlation threshold, generate optimization instructions to iteratively adjust the amplitude / phase spectrum mask of the coded pattern, re-collect and verify until the threshold is met, obtain and lock the current coded pattern parameters and the measured matrix; Step S4: Remove the removable reflector (20) and project the optimized coded pattern onto the sample (19). Collect the signal light returned by the sample (19) and pass it through the time stretching component and photoelectric conversion module in sequence to complete the time stretching and photoelectric conversion. Use the locked measured matrix to recover the image of the sample (19) through the compressed sensing reconstruction algorithm.

Images