High-brightness X-ray light source coupled perovskite imaging system and method

Abstract

The invention discloses a perovskite imaging system and method based on high-brightness X-ray light source coupling. The system sequentially comprises a high-performance X-ray radiation source, an optical divergence element, a coding aperture, an object plane, a detector array based on a perovskite material and a processing unit according to a radiation propagation sequence, emitting X-ray radiation by an X-ray radiation source; intercepting X-ray radiation by using an optical diverging element, and generating a diverging beam emitted from the virtual focus; modulating the divergent beam through the coding aperture and irradiating the divergent beam to an object to be imaged; capturing radiation interacted with the object to be imaged by using a detector array to generate a measured value; the processing unit solves the inverse problem of reversely deducing real physical information of the to-be-imaged object from the coding measurement value according to a preset coding aperture pattern, a system forward model and the measurement value, and reconstructs a high-resolution and high-fidelity image of the to-be-imaged object; the imaging effects of high resolution, high sensitivity, low radiation dose and high fidelity can be realized.

Description

Technical Field

[0001] This invention belongs to the field of perovskite imaging technology, and more specifically, relates to a perovskite imaging system and method based on high-brightness X-ray source coupling. Background Technology

[0002] In the pursuit of high-resolution X-ray imaging, existing technologies face multiple fundamental challenges that are mutually constraining. These challenges stem from a synergistic imbalance between the physical characteristics of radiation sources, limitations of optical systems, and constraints on detector performance. First, real-world X-ray sources have finite physical dimensions, making ideal point source emission impossible. This results in the "penumbra effect," a major cause of blurred image edges, severely limiting image clarity. To mitigate this problem, existing technologies typically employ microfocus X-ray tubes or physical pinhole structures. However, these solutions significantly reduce radiation flux, requiring longer exposure times for imaging. This presents significant limitations in radiation dose-sensitive scenarios such as medical imaging, failing to balance resolution and radiation safety.

[0003] Secondly, while traditional coded aperture imaging technology can theoretically improve the signal-to-noise ratio of images, its design is based on a series of idealized, non-physical assumptions, failing to fully consider the penetrating characteristics of high-energy X-rays and the complex physical environment in practical applications. When applied to high-brightness X-ray imaging scenarios, the deviation between ideal assumptions and actual physical conditions can lead to severe artifacts and systematic noise in the reconstructed images, significantly reducing image fidelity.

[0004] Furthermore, detector performance is one of the core factors limiting the quality of X-ray imaging. Traditional detector materials (such as amorphous selenium α-Se) have reached physical limitations in key indicators such as sensitivity and minimum detectability limit, making it difficult to meet the requirements of next-generation high-resolution, low-dose imaging.

[0005] In recent years, metal halide perovskite materials, with their high atomic number, excellent carrier mobility-lifetime product, and tunable bandgap, have become an ideal choice for constructing high-performance X-ray detectors, offering new possibilities for overcoming traditional technological limitations. However, in existing technologies, coded aperture design and detector physics research are disconnected, lacking a systematic and collaborative optimization method. No technical solution has yet emerged that uses the specific measurable physical parameters of the perovskite detector (such as noise spectrum, high-brightness X-ray response characteristics, and pixel size) as direct input to algorithmically generate customized optimal coded aperture patterns. This design disconnect between components prevents the full realization of the high-performance potential of perovskite detectors, becoming a core technological bottleneck restricting the overall performance improvement of imaging systems.

[0006] In summary, existing X-ray imaging technologies suffer from problems such as the contradiction between resolution improvement and radiation dose control, mismatch between coded aperture design and actual physical conditions, and insufficient synergy between perovskite detectors and other system components. There is an urgent need for an integrated collaborative design scheme to achieve deep integration of radiation source, optical system, coded aperture and perovskite detector, thereby breaking through the performance limitations of existing technologies. Summary of the Invention

[0007] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a perovskite imaging system and method based on high-brightness X-ray source coupling. By constructing an integrated collaborative design system centered on the measured physical characteristics of the perovskite detector, it deeply integrates the flux-loss-free illumination design of sub-micron virtual focal points, the customized generation method of coded aperture based on maximizing mutual information, and the high sensitivity and high resolution characteristics of the perovskite detector. This effectively suppresses penumbra blur in traditional imaging, solves problems such as insufficient detector performance due to component design disconnects, and artifacts and noise interference in high-brightness scenes, and achieves superior performance. The invention achieves synergistic optimization of high spatial resolution (3 lp / mm, with some embodiments reaching 5.0 lp / mm to 10.6 lp / mm) and low radiation dose, while ensuring system stability and high image fidelity in high-throughput imaging scenarios. This significantly improves the overall performance and application adaptability of X-ray imaging, providing a better solution for fields such as medical diagnostics and industrial Micro-CT. The invention solves the technical problem in the prior art that it is difficult to synergistically optimize imaging resolution, sensitivity, and radiation dose, and that the performance of perovskite detectors cannot be fully utilized, thus achieving high-resolution, high-sensitivity, low-radiation-dose, and high-fidelity imaging effects.

[0008] To achieve the above objectives, one aspect of the present invention provides a perovskite imaging system based on high-brightness X-ray source coupling, which includes, in the order of radiation propagation, a high-performance X-ray radiation source, an optical diverging element, an encoding aperture, an object plane, a detector array based on perovskite material, and a processing unit. The optical diverging element is located downstream of the X-ray radiation source and is used to intercept the radiation emitted by the X-ray radiation source and generate a diverging beam emitted from the virtual focal point. The coded aperture is located downstream of the optical diverging element, and its pattern is generated through a physically driven, detector-centric co-design method. The object plane is used to place the object to be imaged; The perovskite-based detector array is used to capture radiation after interacting with the object to be imaged and generate a measurement signal. The processing unit is connected to the perovskite-based detector array and is used to reconstruct an image of the object to be imaged based on the measurements of the perovskite-based detector array.

[0009] Furthermore, the optical diverging element is a multi-capillary focusing lens, the effective size of the virtual focal point is smaller than the physical size of the X-ray radiation source, and the size and position of the virtual focal point are optimized according to the high resolution potential of the detector array.

[0010] Furthermore, the encoding aperture is a single-exposure mask with a QR code MURA array or a checkerboard pattern; the material of the mask includes a high-melting-point metal and a high-thermal-conductivity substrate.

[0011] Furthermore, the collaborative design method takes the measured physical parameters of the detector array as input and aims to maximize the amount of information captured by the imaging system of the object to be imaged. The measured physical parameters of the detector array include at least one of the following: dark current characteristics, shot noise characteristics, thermal noise characteristics, quantum efficiency, and response characteristics to high-brightness X-rays.

[0012] Furthermore, the collaborative design method specifically includes the following steps: S1. Construct a high-fidelity forward model to describe the entire physical process from the object to be imaged to the measurement signal generated on the perovskite detector array. S2. Define the information theory optimization objective and use mutual information as the optimization objective function so that, under the premise that the physical characteristics of the detector array are fixed, the maximum amount of information about the unknown object can be extracted from the measurement value. S3. Execute the optimization algorithm: Solve the optimization objective function, i.e. the maximum amount of information about the unknown object to be imaged, through a numerical optimization algorithm, and finally generate the optimal coded aperture pattern tailored to the specific perovskite detector array.

[0013] Furthermore, the high-fidelity forward model described in step S1 covers a quantitative description of three key physical aspects, including: a complete characterization of the illumination field generated in the virtual focus stage, a precise quantization of the modulation effect of the coded aperture, and a comprehensive integration of the physical characteristics of the perovskite detector array. The high-fidelity forward model described in step S1 also includes a noise model that describes the sources, characteristics, and quantification rules of various noises during the process of the perovskite detector array capturing X-ray radiation and converting it into a measurement signal; the noise model includes dark current noise, shot noise, thermal noise, and high-brightness X-ray response noise. The high-fidelity forward model also includes the thermal load characteristics generated by the virtual focus on the encoding aperture.

[0014] Furthermore, the objective function expression in step S2 is: (1) in, This represents the optimal coded aperture pattern obtained from the final solution; It is a commonly used operator in mathematical optimization, meaning "in the encoding aperture parameter". Within the feasible range, find the one that maximizes the subsequent function value. ”; It refers to mutual information in information theory; Represents the true information of the unknown object to be imaged; Represents the coded measurements captured by the perovskite detector; Assuming the imaging system is a linear model and that the information and noise of the object to be imaged both follow a Gaussian distribution, mutual information... The specific parsing expression is: (2) in: For the imaging system's transmission matrix; Here is the noise covariance matrix; Let the prior covariance matrix of the object be denoted as . It is the identity matrix; This refers to the determinant operation of a matrix.

[0015] Furthermore, the numerical optimization algorithm mentioned in step S3 includes simulated annealing algorithm or genetic algorithm; when designing the mask in step S3, if the optimization algorithm detects that "multi-frame phase hopping" will produce artifacts exceeding the threshold due to lag, it will automatically select a mask of the "single exposure" type. In step S3, the numerical optimization algorithm is solved under the Nyquist theorem constraint, the expression of which is: ;in, Mask cell size for encoding aperture The projection size onto the pixels of the detector array after system optical magnification; The pixel size of the detector array; In step S3, the numerical optimization algorithm is also subject to thermal stability constraints and coded aperture structure constraints; the thermal stability constraint is: the highest temperature of the coded aperture pattern under the action of the virtual focal illumination field is less than or equal to the thermal stability limit temperature of the coded aperture material. The constraints of the coded aperture structure are as follows: the aperture ratio of the coded aperture pattern is greater than 0 and less than the upper limit of the aperture ratio in engineering. At the same time, the unit size of the coded aperture pattern is between the processing accuracy limits of the unit size.

[0016] Furthermore, the reconstruction algorithm employed by the processing unit includes the maximum likelihood expectation maximization algorithm, the compressed sensing-based algorithm, or the deep learning-based algorithm; the spatial resolution of the imaging system is better than 3 lp / mm.

[0017] A second aspect of the present invention provides a perovskite imaging method based on high-brightness X-ray source coupling, implemented using the aforementioned perovskite imaging system based on high-brightness X-ray source coupling, comprising the following steps: S100: Emits X-ray radiation through a high-performance X-ray radiation source; S200. The X-ray radiation is intercepted using an optical diverging element to generate a diverging beam emitted from the virtual focal point 125; S300. The divergent beam is modulated by the coded aperture so that the modulated beam illuminates the object to be imaged. S400: Uses a detector array based on perovskite material to capture radiation after it interacts with the object to be imaged, and generates encoded measurement values. S500: The processing unit uses a reconstruction algorithm to solve the inverse problem of "inferring the true information of the object from the measured value" based on the preset coded aperture pattern, the system forward model and the coded measurement value, and outputs a high-resolution, high-fidelity image of the object to be imaged.

[0018] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: (1) The perovskite imaging method and system based on high-brightness X-ray source coupling of the present invention can realize the collaborative adaptation of system-level components and fully release the performance potential of perovskite detectors. Taking the measured physical characteristics of perovskite detectors as the core, the size and position of virtual focus are precisely optimized and matched with the high-resolution potential of detectors (such as pixel size and target resolution). The coded aperture pattern adapted to the characteristics of detectors is generated through a physical-driven collaborative design method, so that the design of optical front end (virtual focus, coded aperture) fully serves the maximization of detector advantages. This solves the problem of detector performance not being released due to component design disconnect in the prior art and ensures the improvement of imaging quality at the system level.

[0019] (2) The perovskite imaging method and system based on high-brightness X-ray source coupling of the present invention can significantly improve imaging resolution and radiation dose efficiency; the submicron virtual focus constructed by the optical diverging element not only suppresses penumbra blur in traditional technology, but also endows the radiation field with high spatial coherence, providing a physical basis for high contrast modulation of the coded aperture; at the same time, the flux-free characteristics of the virtual focus are matched with the high sensitivity of the perovskite detector, and combined with the low photon loss design of the coded aperture, the present invention can achieve a spatial resolution better than 3lp / mm while controlling the radiation dose at a level far lower than that of traditional imaging technology (such as the radiation dose in medical scenarios can be reduced to 0.1mSv), and can achieve synergistic optimization of high resolution and low dose.

[0020] (3) The perovskite imaging method and system based on high-brightness X-ray source coupling of the present invention can ensure imaging fidelity and system stability in high-brightness scenes; the high-fidelity forward model, through precise quantitative mathematical form, fully incorporates the characteristics and mechanisms of the three core physical links in the entire imaging chain, namely: the spatial distribution, intensity law and coherence characteristics of the illumination field formed by the virtual focus stage, the intensity / phase modulation law of the encoded aperture to the incident radiation (related to the encoded pattern and material parameters), and the photoelectric conversion characteristics, noise characteristics and response law of the perovskite detector array under high-brightness X-rays, realizing The high-fidelity forward model provides a faithful mapping of the entire physical process from radiation irradiation of an object to the output measurement signal from the detector. It also incorporates the thermal load characteristics of the coded aperture and, combined with the selection of coded aperture materials with high melting points and high thermal conductivity, ensures the thermal stability of the system under high-brightness X-ray flux. At the same time, the time response defects of perovskite detectors are avoided through co-design, and the sampling and adaptation of coded patterns are completed under the constraints of the Nyquist theorem. This effectively eliminates artifacts, aliasing, and noise interference in high-brightness scenes, enabling the reconstructed image to have a high signal-to-noise ratio and high fidelity, and making it stable for high-throughput imaging scenarios such as industrial Micro-CT.

[0021] (4) The perovskite imaging method and system based on high-brightness X-ray source coupling of the present invention transforms the design of the coded aperture into an information theory optimization problem with the goal of maximizing mutual information. By quantifying the entire physical process through a high-fidelity forward model, the mask design is upgraded from traditional heuristic guessing to a calculable and quantifiable scientific method. The generated coded aperture pattern not only meets engineering processing constraints (such as unit size and aperture ratio) but also maximizes information capture efficiency in a real physical environment, thereby improving the practicality and industrialization value of the technical solution. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the perovskite imaging system based on high-brightness X-ray source coupling according to an embodiment of the present invention; Figure 2 A schematic diagram of setting up ray tracing for traditional magnified imaging; Figure 3 This is a schematic diagram of ray tracing according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure when the encoding aperture is a two-dimensional MURA array in an embodiment of the present invention; Figure 5 This is an enlarged schematic diagram of the cell structure when the encoding aperture is a two-dimensional MURA array in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the design principle of the mask unit size in an embodiment of the present invention; Figure 7 This is a schematic flowchart of a perovskite imaging method based on high-brightness X-ray source coupling according to an embodiment of the present invention.

[0023] In all the accompanying drawings, the same reference numerals denote the same technical features, specifically: 110-X-ray radiation source, 120-optical diverging element, 125-virtual focal spot, 130-coded aperture, 140-object plane, 145-imaging object, 150-detector array based on perovskite material, and 160-processing unit. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0025] like Figure 1 As shown, one aspect of the present invention provides a perovskite imaging system based on high-brightness X-ray source coupling, which includes, in the order of radiation propagation: a high-performance X-ray radiation source 110, an optical diverging element 120, an encoding aperture 130, an object plane 140, a detector array 150 based on perovskite material, and a processing unit 160. The optical diverging element 120 is located downstream of the X-ray radiation source 110 and is used to intercept the radiation emitted by the X-ray radiation source 110 and generate a diverging beam emitted from the virtual focus 125. The coded aperture 130 is located downstream of the optical diverging element 120, and its pattern is generated by a physically driven, detector-centric co-design method. The object plane 140 is used to place the object to be imaged 145; The perovskite-based detector array 150 is used to capture radiation after interacting with the object to be imaged 145 and generate a measurement signal. The processing unit 160 is connected to the perovskite-based detector array 150 and is used to reconstruct an image of the object 145 to be imaged based on the measurement values ​​of the perovskite-based detector array 150. When the system is in operation, the radiation emitted by the high-performance X-ray radiation source 110 is illuminated by the optical diverging element 120 to form a virtual focal point, and then modulated by the coded aperture 130 customized for the perovskite detector before irradiating the object 145. The radiation that interacts with the object is captured by the perovskite detector array 150 to form coded measurement values. The processing unit 160 receives these measurement values ​​and uses precise knowledge about the pattern of the coded aperture 130 and the forward model of the system to calculate and reconstruct a high-fidelity image of the object 145 by solving an inverse problem.

[0026] Furthermore, the core quantitative characterization of the high-brightness X-rays described in this invention is photon flux density, corresponding to the following key operating parameter ranges for the radiation source: Tube voltage: 80-120kV for medical diagnostic scenarios, 150-200kV for industrial Micro-CT scenarios; Tube current: 50-100mA in medical diagnostic scenarios, 200-300mA in industrial Micro-CT scenarios; Photon flux density: ≥10 10 Photons / (s·mm 2 (Measured value at the detector receiving surface) to ensure that the number of photons passing through the coded aperture and the detector per unit time meets the imaging requirements of "low exposure time + high signal-to-noise ratio"; lower limit: based on "being able to match the high sensitivity of the perovskite detector and generate an effective measurement signal within an exposure time of 50-300ms", to avoid the signal-to-noise ratio being too low and the loss of image details due to insufficient photon flux. The lower limit of the key operating parameters of the corresponding radiation source is based on "being able to match the high sensitivity of the perovskite detector and generate an effective measurement signal within an exposure time of 50-300ms", so as to avoid the signal-to-noise ratio being too low and the loss of image details due to insufficient photon flux. Upper limit: The constraint is that "after the submicron virtual focus is formed by the optical diverging element, the thermal load of the encoding aperture does not exceed its thermal stability limit (e.g., the maximum temperature of the tungsten material mask is ≤800℃)" to prevent the encoding aperture from deforming and the detector from saturating due to excessive photon flux. Further, please refer to Figure 2 A schematic diagram of ray tracing in traditional magnified imaging setups, illustrating the formation process of significant penumbra blur due to the finite size of the radiation source (umbra region 240, penumbra region 250); please refer to... Figure 3 The schematic diagram of the ray tracing of the present invention shows that a virtual focus 125 is created by the optical diverging element 120, thereby greatly suppressing the penumbra effect on the object plane 140. Furthermore, the optical diverging element 120 is a multi-capillary focusing lens, and the effective size of the virtual focus 125 is smaller than the physical size of the X-ray radiation source 110. The size and position of the virtual focus 125 are optimized according to the high-resolution potential of the detector array 150. Specifically, this invention addresses the penumbra problem in traditional imaging technology by inserting an optical diverging element 120 (such as a multi-capillary focusing lens) between the X-ray radiation source 110 and the object to be imaged 145. The optical diverging element 120 can cause the radiated light to diverge, forming a radiation propagation effect that appears to originate from the virtual focus 125 in front of it. By precisely designing the structural parameters of the optical diverging element 120, a virtual focus 125 with an effective size much smaller than the physical size of the X-ray radiation source 110 can be constructed. This virtual focus 125 can dominate the geometric characteristics of the radiation field irradiating the object to be imaged 145, thereby greatly suppressing the penumbra blurring phenomenon. More importantly, the submicron-scale virtual focus 125 constructed by the optical diverging element 120 is not only the core source of the imaging system's high resolution, but its extremely small source size (s) also enables the radiation source to possess high spatial coherence (coherence length). This high spatial coherence is a key physical prerequisite for the subsequent encoding aperture 130 (especially phase-sensitive masks or MURA arrays) to generate high-contrast interference or modulation patterns, forming the core physical basis for computational imaging in this invention. Furthermore, the design parameters such as the size and position of the virtual focus 125 are precisely optimized based on the high-resolution potential (e.g., pixel size, target resolution) of the perovskite-based detector array 150, thus providing optimal illumination conditions to fully utilize the high-resolution performance of the perovskite-based detector array 150. Simultaneously, the illumination scheme has no radiation flux loss, perfectly matching the high sensitivity characteristics of the perovskite-based detector array 150, ultimately achieving high-resolution imaging with low radiation dose.

[0027] Furthermore, the encoding aperture 130 is a single-exposure mask with a two-dimensional MURA array or a checkerboard pattern; the material of the mask includes a high-melting-point metal and a high-thermal-conductivity substrate, wherein the high-melting-point metal is tungsten and the high-thermal-conductivity substrate is diamond.

[0028] Furthermore, unlike existing technologies that employ general patterns based on purely mathematical ideal models, this invention proposes a physical-driven, detector-centric co-design method for the design of the coded aperture 130. This physical-driven, detector-centric co-design method transforms the design of the coded aperture 130 into an information-theoretic optimization problem, the goal of which is to maximize the amount of information captured by the entire imaging system (from the mask to the perovskite detector) about the object to be imaged 145; that is, the pattern of the coded aperture 130 is generated through a physical-driven co-design method. The collaborative design method takes the measured physical parameters of the detector array 150 as input and aims to maximize the amount of information captured by the imaging system of the object 145 to be imaged; the measured physical parameters of the detector array include at least one of dark current characteristics, shot noise characteristics, thermal noise characteristics, quantum efficiency and response characteristics to high-brightness X-rays. The collaborative design method specifically includes the following steps: S1. Construct a high-fidelity forward model to describe the entire physical process from the object to be imaged 145 to the measurement signal generated on the perovskite detector array 150. S2. Define the information theory optimization objective and use mutual information as the optimization objective function so that, under the premise that the physical characteristics of the detector array are fixed, the maximum amount of information about the unknown object can be extracted from the measurement value. S3. Execute optimization algorithm: Solve the optimization objective function through numerical optimization algorithm, that is, the maximum amount of information about the unknown object to be imaged 145, and finally generate the optimal coded aperture pattern tailored to the specific perovskite detector array 150. Furthermore, the high-fidelity forward model described in step S1 is a mathematical model capable of accurately reproducing the entire physical process from X-ray radiation irradiation of the object to be imaged to the output measurement signal from the perovskite detector array. Its core encompasses the quantitative description of three key physical links, ensuring a faithful mapping of the model to the actual imaging process, as follows: A complete characterization of the "illumination field generated in the virtual focus stage": The high-fidelity forward model will accurately describe the physical characteristics of the radiation emitted by the virtual focus 125 formed by the optical diverging element 120 (such as a multi-capillary focusing lens), including but not limited to the spatial distribution of the illumination field (such as Gaussian distribution, uniform distribution, etc.), the radiation intensity distribution law, spatial coherence (coherence length characteristics determined by the size of the virtual focus), propagation direction and divergence angle, as well as the influence of the size and position parameters of the virtual focus on the entire illumination field, fully reproducing a high-resolution illumination scene without flux loss; Precise quantification of the modulation effect of the coded aperture: The model will incorporate in detail the physical modulation process of the coded aperture (such as MURA array, checkerboard pattern mask) on the incident radiation, including the modulation rules corresponding to parameters such as the unit size, aperture ratio, spatial arrangement pattern, and material properties (such as the absorption rate of the light-shielding layer and the transmittance of the substrate). Specifically, it is reflected in the spatial distribution changes of the intensity and phase (for phase-sensitive masks) of the radiation after passing through the coded aperture, as well as the screening and encoding mechanism of the coded aperture on the radiation. A comprehensive integration of the physical characteristics of perovskite detector arrays: The model incorporates the core physical characteristics of perovskite detector arrays in the process of receiving radiation and converting it into electrical signals. This includes the detector's basic photoelectric conversion characteristics (such as quantum efficiency, carrier mobility-lifetime product, response time, pixel size, and array scale), as well as noise characteristics constructed based on the detector's measured parameters (such as dark current noise, shot noise, thermal noise, and response noise under high-brightness X-rays). It also covers the detector's response to radiation (such as the linear / nonlinear relationship between input radiation intensity and output electrical signal, saturation characteristics under high-brightness radiation, etc.), thus fully reproducing the physical process by which the detector converts X-ray signals into measurable electrical signals. The high-fidelity forward model integrates the physical mechanisms and quantitative parameters of three core components: "generation and propagation of virtual focal illumination field", "modulation and coding of radiation by coded aperture", and "signal conversion and noise introduction of perovskite detector", to achieve accurate mathematical modeling of the entire imaging physical process. Furthermore, the high-fidelity forward model described in step S1 also includes a noise model that describes the sources, characteristics, and quantification laws of various noises during the process of the perovskite detector array capturing X-ray radiation and converting it into a measurement signal. The core function of the noise model is to abandon the idealized assumption of "uniform Gaussian noise" in traditional imaging technology, accurately reproduce the real noise behavior based on the actual measured physical parameters of the detector, and ensure that the forward model's prediction of the measurement signal is highly consistent with the actual physical process. The noise model is no longer a simple uniform Gaussian noise assumption, but is constructed based on the physical parameters actually measured for the selected perovskite detector, including dark current noise, shot noise, thermal noise, and high-brightness X-ray response noise. The noise model of this invention adopts a "superposition model" to integrate the above four types of noise components, that is, the variance of the total noise is the sum of the variances of each independent noise component. Among them, the variance of each noise component is a quantization function based on the measured parameters of the detector, rather than a fixed constant, to ensure that the model can accurately adapt to the individual characteristics of the target perovskite detector and avoid the failure of the coded aperture design optimization or the appearance of artifacts in the reconstructed image due to noise modeling deviation. The noise model provides realistic constraints for the co-design of the coded aperture: when the optimization algorithm solves for the optimal coded aperture pattern, the noise model can accurately quantify the signal-to-noise ratio (SNR) changes under different coded patterns, ensuring that the optimization results (such as aperture ratio, cell size, and material selection) can maximize information capture in a real noise environment; the noise model can improve the fidelity of image reconstruction: when the processing unit 160 executes reconstruction algorithms (such as MLEM and compressed sensing), it can perform noise suppression and error correction based on the noise model, avoiding misjudging noise as object signals, thereby reducing artifacts in the reconstructed image; the noise model can match the characteristic advantages of perovskite detectors: by accurately modeling the low-noise characteristics of high-sensitivity perovskite detectors (such as low dark current and low shot noise due to high quantum efficiency), it ensures that the coded aperture design can fully utilize this advantage to achieve high signal-to-noise ratio imaging at low doses.

[0029] Furthermore, the high-fidelity forward model constructed in this invention, in addition to the virtual focus illumination field characteristics, coded aperture modulation effect, and perovskite detector physical characteristics already covered, also includes a quantitative description of the physical phenomenon of "X-ray radiation induced by the virtual focus generating a thermal load on the coded aperture." That is, it accurately characterizes the generation mechanism, distribution law, and influencing parameters of this thermal load through mathematical form, enabling the model to fully reproduce the real physical environment under high-brightness X-ray imaging scenarios.

[0030] Furthermore, the ultimate goal of designing the coded aperture 130 in step S2 is to find an optimal coded aperture pattern that, given the physical characteristics of the perovskite detector array, can extract the maximum amount of information about the unknown object to be imaged 145 from the measurements; this scheme uses mutual information from information theory as the optimization objective function. The objective function expression is: (1) in, This represents the optimal encoded aperture pattern obtained from the final solution (such as the parameter combination of a MURA array or a checkerboard mask, including cell arrangement, aperture ratio, cell size, etc.). It is a commonly used operator in mathematical optimization, meaning "in the encoding aperture parameter". Within the feasible range, find the one that maximizes the subsequent function value. ”; here “ "These are the encoding aperture parameters to be optimized (such as unit size, aperture position, and other variables); It refers to mutual information in information theory, used to quantify the "true information of the object to be imaged". "and the measured value output by the detector" The degree of information correlation between them: Represents the true information of the unknown object to be imaged (such as the object's internal structure, density distribution, geometric shape, and other physical characteristics). Represents the coded measurements captured by the perovskite detector (i.e., the electrical signal output by the detector after X-rays are modulated by the coded aperture and interact with the object, containing effective coded information and noise); mutual information. The larger the value, the higher the measured value. The real information of the objects contained The more, the more from The higher the fidelity of the reconstructed object image; Formula (1) means: by adjusting the encoding aperture parameter Find the information of the object to be imaged. "and detector measurements" Mutual information between Largest coded aperture pattern ,Right now It is the optimal coding aperture for "extracting the most object information from the measurement value"; Assuming the imaging system is a linear model (i.e., the system's input and output satisfy a linear superposition relationship) and that the information and noise of the object to be imaged both follow a Gaussian distribution, mutual information... The specific parsing expression is: (2) in: For the imaging system's transmission matrix, coded aperture parameters are used. The decision describes the linear transformation process from "object information to detector measurements"; The noise covariance matrix characterizes the statistical properties of noise in perovskite detectors (such as the distribution and correlation of dark current and shot noise). The prior covariance matrix of the object represents the statistical distribution characteristics of the information of the object to be imaged (such as the spatial correlation of the internal structure of the object). It is the identity matrix, used to maintain dimensional consistency in matrix operations; This is a matrix determinant operation used to convert matrix information into scalar values, ultimately yielding the mutual information value. This analytical formula (2) allows for the quantitative calculation of specific coding aperture parameters. The corresponding mutual information value provides a computable mathematical tool for "finding the optimal coding aperture that maximizes mutual information". In optimization algorithms (such as simulated annealing and genetic algorithms), the mutual information can be calculated by substituting into the formula for each candidate coding aperture A in each iteration, thereby selecting the optimal solution. Furthermore, the collaborative optimization method of this invention can automatically and quantitatively calculate the optimal mask pattern that suits the characteristics of any perovskite detector, transforming the previously experience-based mask design into a quantitative and computationally comprehensible scientific method. The key to achieving this lies in the forward model in step S1 and the optimization algorithm in step S3, which transforms the physical characteristics of the perovskite detector into three core constraints / synergistic advantages: specifically including: Constraint 1: Detector time response (image lag) Perovskite detectors exhibit image lag (inter-frame signal residue) due to ion migration and deep-level traps. The forward model in step S1 of this invention quantitatively describes this characteristic. The numerical optimization algorithm in step S3 includes simulated annealing or a genetic algorithm. When designing the mask, if the optimization algorithm in step S3 detects that "multi-frame phase pacing" will produce artifacts exceeding the threshold due to lag, it will automatically select a "single-exposure" type mask (such as a two-dimensional MURA or a checkerboard pattern). This single-image method physically avoids the detector's time response defects, ensuring that the encoding process is not interfered with by inter-frame residue. The encoding aperture is a single-exposure mask with a two-dimensional MURA array or a checkerboard pattern (such as...). Figure 4 and Figure 5 (as shown); the material of the photomask includes a high-melting-point metal and a high-thermal-conductivity substrate, wherein the high-melting-point metal is tungsten and the high-thermal-conductivity substrate is diamond; Constraint 2: Detector pixel sampling (Nyquist theorem) Physical principle: In step S3, the numerical optimization algorithm is solved under the constraint of the Nyquist theorem; the Nyquist theorem requires that "the sampling frequency must be greater than twice the signal frequency", otherwise undersampling (information loss, aliasing artifacts) will occur; corresponding to the imaging system, the projection size of the mask unit must be greater than twice the size of the detector pixel; The model's role: In step S1, the forward model accurately calculates the "mask element size". After optical magnification by the system, at the detector pixel Projection size "(like Figure 6 (as shown) Collaborative design result: In a specific embodiment of the present invention, the Nyquist theorem constraint constrains the projection size of the mask unit for encoding aperture onto the detector array pixels after system amplification. satisfy ,in, Mask cell size for encoding aperture The projection size onto the pixels of the detector array after system optical magnification; The pixel size of the detector array is used; the Nyquist sampling constraint ensures that the modulation pattern of the coded aperture can be correctly sampled by the pixel array of the perovskite detector, avoiding information loss and aliasing artifacts caused by undersampling. In step S3, the optimization algorithm will " "As a hard constraint, the optimal mask is only searched within the range that meets this condition; this ensures that the modulation pattern generated by the coding aperture 130 can be correctly sampled by the discrete pixel array of the perovskite detector array 150, preventing information loss and aliasing artifacts caused by undersampling, and ensuring the complete transmission of information." Synergistic advantages: High detector sensitivity (high gain) Detector characteristics: Perovskite detectors have the advantages of "high sensitivity and high photoelectric conversion gain"; even with few incident photons, they can output a sufficiently strong signal. The role of the model: In step S1, the forward model incorporates these "high sensitivity and high gain" characteristics (such as quantum efficiency and signal amplification factor) into the calculation; Collaborative design results: This advantage allows for greater design flexibility in the optimization algorithm during step S3: the algorithm can select a mask with a lower aperture ratio (e.g., 50%); such a mask will block some photons, but it can improve the robustness (anti-interference) and contrast of the encoding; and the high sensitivity of the perovskite detector can compensate for this photon loss, allowing the system to maintain a high signal-to-noise ratio even at low radiation doses; ultimately, it achieves dual optimization of "dose efficiency (low dose)" and "encoding efficiency (high robustness)," fully leveraging the advantages of the perovskite detector.

[0031] Furthermore, the optimization objective of this invention is to "maximize mutual information". "To find the optimal coded aperture pattern" However, this optimization process must also be subject to the following two types of constraints: Thermal stability constraint: The highest temperature of the coded aperture pattern under the action of the virtual focal illumination field (calculated from the thermal load characteristics in the forward model) is less than or equal to the thermal stability limit temperature of the coded aperture material (a known constant, such as the safe operating temperature of tungsten ≤800℃, determined by the material characteristics of the mask); thus ensuring that the coded aperture does not deform or get damaged under the thermal load of high-brightness X-rays, and ensuring the stable operation of the system at high throughput; Encoded aperture structure constraints: The aperture ratio of the encoded aperture pattern is greater than 0 and less than the engineering upper limit of the aperture ratio. At the same time, the unit size of the encoded aperture pattern is between the processing accuracy limits of the unit size. This ensures that the aperture ratio and unit size of the encoded aperture meet the requirements of engineering processing accuracy and structural strength, and is feasible for actual production. In the collaborative design of the coded aperture pattern of the present invention, under the premise of meeting engineering constraints such as "sampling without aliasing, thermal stability of the mask, and fabrication of the structure", the mutual information between "real information of the object" and "detector measurement value" is maximized by adjusting the parameters of the coded aperture pattern (cell arrangement, aperture ratio, size, etc.), and the optimal coded aperture pattern is finally determined.

[0032] Furthermore, the reconstruction algorithm employed by the processing unit includes the maximum likelihood expectation maximization algorithm, the compressed sensing-based algorithm, or the deep learning-based algorithm, and the spatial resolution of the imaging system is better than 3 lp / mm.

[0033] like Figure 7 As shown, a second aspect of the present invention provides a perovskite imaging method based on high-brightness X-ray source coupling, comprising the following steps: S100, emits X-ray radiation through a high-performance X-ray radiation source 110; Specifically, this includes: activating the high-performance X-ray radiation source 110 of the present invention, so that it stably outputs X-ray radiation according to preset working parameters (such as tube voltage, tube current, radiation intensity, etc., adapted to specific application scenarios such as medical diagnosis and industrial testing); S200: The X-ray radiation is intercepted by the optical diverging element 120 to generate a diverging beam emitted from the virtual focus 125; Specifically, this includes: placing an optical diverging element 120 (such as a multi-capillary focusing lens) in the downstream radiation propagation path of the X-ray radiation source 110; when the X-ray radiation emitted by S100 irradiates the optical diverging element 120, the element will diverge and control the radiation light, so that the light originally emitted from the physical radiation source will present a propagation effect of "emitting from the virtual focus 125 in front of the element" after being controlled; by precisely designing the structural parameters of the optical diverging element 120, the effective size of the virtual focus 125 is much smaller than the physical size of the X-ray radiation source 110, and it has high spatial coherence, which can greatly suppress penumbra blur in subsequent imaging and provide the necessary physical premise for the modulation of the coding aperture; S300, The divergent beam is modulated by the coding aperture 130 so that the modulated beam illuminates the object to be imaged 145. Specifically, this includes: setting an coded aperture 130 downstream of the optical diverging element 120 (the pattern of this coded aperture is tailored based on the measured physical parameters of the perovskite detector array 150 through information theory optimization algorithms, such as a two-dimensional MURA array, a checkerboard pattern, etc.); when the diverging beam generated by S200 illuminates the coded aperture 130, the coded aperture will modulate the intensity or phase of the beam according to its own pattern structure (i.e., "encode" the beam), so that the beam carries specific coded information; the modulated beam will continue to travel along the propagation path and finally uniformly illuminate the object to be imaged 145 placed on the object plane 140, completing the preparation for the interaction between the coded beam and the object; S400: The perovskite-based detector array 150 captures the radiation that interacts with the object to be imaged 145 and generates coded measurements. Specifically, this includes: after the object to be imaged 145 is irradiated by the modulated light beam, the light beam interacts with the object's internal structure, density distribution, and other physical characteristics (such as absorption, scattering, and transmission); a perovskite-based detector array 150 is positioned downstream of the object to be imaged 145 to capture the X-ray radiation after the interaction with the object. The perovskite detector array, with its high atomic number and excellent carrier mobility-lifetime product, efficiently converts the captured radiation signal into an electrical signal while retaining the encoded information carried by the light beam, ultimately generating an encoded measurement value (this measurement value contains the physical characteristic information and encoded information of the object to be imaged, and also incorporates the detector's true noise characteristics). S500: The processing unit 160 uses a reconstruction algorithm to solve the inverse problem of "inferring the true information of the object from the measured value" based on the preset coded aperture 130 pattern, the system forward model and the coded measurement value, and outputs a high-resolution, high-fidelity image of the object 145 to be imaged. Specifically, the processing unit 160 pre-stores customized pattern data of the coded aperture 130, as well as a high-fidelity system forward model constructed in this invention (this model fully characterizes the entire physical process from X-ray emission, virtual focus generation, coded modulation to detector capture, including illumination field characteristics, coded modulation rules, detector physical characteristics, and thermal load characteristics, etc.); the processing unit 160 receives the coded measurement values ​​generated by S400, and using the coded aperture pattern and the system forward model as known conditions, solves the inverse problem of "inferring the true physical information of the object to be imaged from the coded measurement values" by adopting an adapted reconstruction algorithm, completes the decoding, noise suppression, and signal restoration of the measurement values, and finally reconstructs a high-fidelity image that can truly reflect the internal structure, physical characteristics (density distribution and geometric shape) of the object to be imaged 145, achieving high-resolution, high-sensitivity, and low-noise imaging targets; the reconstruction algorithm includes the maximum likelihood expectation maximization algorithm, the compressed sensing-based algorithm, or the deep learning algorithm. This invention achieves unprecedented image fidelity, spatial resolution, and detection sensitivity by using a perovskite detector as the core and integrating the X-ray source, optical front end, coding aperture, and detector into a cohesive design.

[0034] Furthermore, in step S300, the encoded aperture 130 is generated by a numerical optimization algorithm based on the measured physical parameters of the detector array 150, and the size and position of the virtual focus 125 are optimized according to the high resolution potential of the detector array 150.

[0035] The effects of the present invention will now be described with reference to specific embodiments.

[0036] Example 1: High-resolution, low-dose imaging device for medical diagnosis Component selection: High-performance X-ray radiation source: a microfocus X-ray tube with a tube voltage of 80-120kV and a tube current of 50-100mA is selected, and the physical source size is 50μm; Optical divergence element: A multi-capillary focusing lens is used, with an aperture size of 10μm and a length of 50mm. Through optical design, the virtual focal point size is 0.8μm and is located 100mm in front of the multi-capillary focusing lens. Perovskite detector array: Formamidinium lead iodide (FAPbI3) perovskite material is used, with a pixel size of 50 μm and an array size of 2048 × 2048. The measured dark current is 10. -12 A / cm 2 The shot noise level is 10. -6 A, Quantum efficiency ≥90% under 100kV X-rays; Encoding aperture: A two-dimensional MURA array mask is generated using a collaborative design method. The mask cell size is 200 μm. The projection size onto the detector pixel after system magnification is... (satisfy (Constrained by the Nyquist theorem); the photomask uses tungsten (melting point 3410℃) as the light-shielding layer and diamond as the substrate (thermal conductivity 2000W / (m・K)), with an aperture ratio of 50%; Processing unit: It adopts an industrial computer equipped with a GPU and has a built-in deep learning-based image reconstruction algorithm.

[0037] Collaborative design process: S1. Construct a high-fidelity forward model: The model incorporates the Gaussian illumination field distribution formed by the virtual focus (0.8μm), the modulation function of the MURA array, and the measured noise model (dark current + shot noise + thermal noise) and high-brightness X-ray response curve of the perovskite detector; at the same time, the thermal load (maximum temperature ≤300℃) generated by the virtual focus on the encoding aperture is considered. S2. Define the optimization objective: The objective function is to maximize the mutual information between the detector measurements and the true information of the object. S3. Optimization Algorithm: A genetic algorithm is used with 1000 iterations and a population size of 50 to finally generate a MURA array pattern adapted to the FAPbI3 detector. Imaging effect: Imaging a human chest phantom with an exposure time of 50ms and a radiation dose of 0.1mSv (far lower than the 1-10mSv of traditional CT), the imaging spatial resolution reaches 6.8lp / mm, the image signal-to-noise ratio is ≥50dB, there are no obvious artifacts, and it can clearly identify microcalcifications with a diameter of 0.5mm, meeting the diagnostic needs of early lung cancer screening.

[0038] Example 2: High-throughput imaging device for industrial Micro-CT Component selection: High-performance X-ray radiation source: A high-power X-ray source with a tube voltage of 150-200kV and a tube current of 200-300mA is selected, and the physical source size is 100μm; Optical divergence element: A multi-capillary focusing lens is used, with an aperture size of 15μm and a length of 80mm. The generated virtual focal point size is 1.2μm and is located 150mm in front of the multi-capillary focusing lens. Perovskite detector array: cesium lead bromide (CsPbBr3) perovskite material is used, with a pixel size of 25 μm and an array size of 4096 × 4096. The measured dark current is 5 × 10⁻⁶. -13 A / cm 2 The response time under 200kV X-rays is ≤10μs; Encoded aperture: A checkerboard pattern single-exposure mask is generated using a collaborative design method. The mask unit size is 100μm. The projection size onto the detector pixel after system magnification is... (satisfy (Constrained by Nyquist's theorem); The mask material is a tungsten-diamond composite structure with an aperture ratio of 60%; Processing unit: Employs a multi-GPU cluster, equipped with the Maximum Likelihood Expectation Maximization (MLEM) reconstruction algorithm, with 50 iterations; Collaborative design process: S1. Construct a high-fidelity forward model: Focus on incorporating the strong illumination field characteristics of high-power X-ray sources, the modulation effect of checkerboard masks, and the linear response characteristics and thermal noise suppression capabilities of CsPbBr3 detectors under high-brightness X-rays; at the same time, quantify the thermal load (maximum temperature ≤400℃) generated by the virtual focus on the encoding aperture to ensure the thermal stability of the mask. S2. Define the optimization objective: with the core objective being to maximize the amount of information captured by the imaging system on the internal defects of industrial parts, mutual information is used as the optimization objective function. S3. Optimization Algorithm: The simulated annealing algorithm is adopted, with an initial temperature of 100 and a cooling rate of 0.95, to finally generate a checkerboard mask pattern adapted to the CsPbBr3 detector. Imaging effect: Micro-CT imaging of aero-engine turbine blades (made of titanium alloy with a thickness of 5 mm) can be performed in 30 seconds per scan (compared to 3-5 minutes for traditional devices). The spatial resolution of the imaging reaches 10.2 lp / mm, which can clearly distinguish tiny pores and cracks with a diameter of 0.1 mm inside the blade. The images are free of thermal noise artifacts and penumbra blur, meeting the non-destructive testing requirements of precision industrial parts.

[0039] This invention enables system-level co-optimization centered on the perovskite detector: all front-end designs (virtual focus, coded aperture) are developed to maximize the unique advantages of the perovskite detector. The size and position of the virtual focus are optimized based on the high-resolution potential of the perovskite detector; the pattern of the coded aperture is calculated directly from the measured physical parameters of the perovskite detector (including noise spectrum, quantum efficiency, and response characteristics to high-brightness X-rays). This co-design ensures optimal matching of the performance of each system component, fundamentally improving the imaging quality of the entire system.

[0040] This invention significantly improves the spatial resolution and sensitivity of imaging devices: by integrating a high-performance X-ray source with a perovskite detector and optimizing the brightness adjustment and coupling method of the source, this invention fully explores and achieves the optimal performance of perovskite imaging devices under different X-ray sources. Combined with the effective suppression of penumbra blur using virtual focus technology and the excellent photoelectric properties of perovskite materials, the system of this invention can achieve a spatial resolution better than 3 lp / mm, and in specific embodiments, even reaches a level of 5.0 lp / mm to 10.6 lp / mm, fully meeting and exceeding the requirements of medical diagnosis.

[0041] This invention effectively improves image quality under high-brightness X-rays: It delves into and utilizes the robust response characteristics of perovskite materials to high-brightness X-rays; by incorporating this characteristic into the design model of the coded aperture and optimizing the operating parameters of the X-ray source, further, the forward model in step S1 of this invention not only considers the detector's response under high brightness but must also include the extreme thermal load generated by the virtual focus 125 (which focuses high-power X-rays to the micrometer level) on the coded aperture 130. Therefore, the optimization algorithm in step S3 searches for the optimal solution A. * When high throughput is required, the system will automatically select a mask scheme with better thermal stability (e.g., high-melting-point materials such as tungsten, or high-thermal-conductivity substrates such as diamond) to ensure that both the front-end optics (mask) and the back-end detector (perovskite) can operate stably at high throughput, thereby obtaining high signal-to-noise ratio, artifact-free images. This enables the system to maintain stable operation in applications requiring high throughput to shorten imaging time or improve signal-to-noise ratio (such as industrial micro-CT), obtaining high signal-to-noise ratio, artifact-free images.

[0042] This invention fully leverages the high-performance advantages of perovskite detectors through deep collaborative design among components, effectively resolving the core contradictions of existing technologies and providing a superior X-ray imaging solution for fields such as medical diagnosis and industrial inspection.

[0043] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A perovskite imaging system based on coupling with high-brightness X-ray sources, characterized by, The high-performance X-ray radiation source (110), the optical divergence element (120), the coded aperture (130), the object plane (140), the perovskite material-based detector array (150) and the processing unit (160) are sequentially arranged in the order of radiation propagation; The optical divergence element (120) is located downstream of the X-ray radiation source (110) and is used to intercept the radiation emitted by the X-ray radiation source (110) and generate a divergent light beam emitted from a virtual focal point (125); The coded aperture (130) is located downstream of the optical divergence element (120), and the pattern thereof is generated by a physically driven, detector-centered collaborative design method; The object plane (140) is used to place the object (145) to be imaged; The perovskite material-based detector array (150) is used to capture the radiation after interacting with the object (145) to be imaged and generate a measurement signal; The processing unit (160) is connected to the perovskite material-based detector array (150) and is used to reconstruct a high-resolution, high-fidelity image of the object (145) to be imaged according to the measurement value of the perovskite material-based detector array (150).

2. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 1, characterized in that, The optical divergence element (120) is a multi-capillary focusing mirror, the effective size of the virtual focal point (125) is smaller than the physical size of the X-ray radiation source (110), and the size and position of the virtual focal point (125) are optimized according to the high-resolution potential of the detector array (150).

3. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 2, characterized in that, The coded aperture (130) is a single-exposure mask plate of a two-dimensional code MURA array or a checkerboard pattern; the material of the mask plate includes a high-melting-point metal and a high-thermal-conductivity substrate.

4. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 3, characterized in that, The collaborative design method takes the measured physical parameters of the detector array (150) as input and aims to maximize the amount of information captured by the imaging system about the object (145) to be imaged. The measured physical parameters of the detector array (150) include at least one of dark current characteristics, shot noise characteristics, thermal noise characteristics, quantum efficiency, and response characteristics to high-intensity X-rays.

5. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 4, characterized in that, The collaborative design method specifically includes the following steps: S1, constructing a high-fidelity forward model to describe the entire physical process of generating a measurement signal from the object (145) to be imaged to the perovskite detector array (150); S2, defining an information theory optimization target, using mutual information as an optimization objective function, so that under the premise of fixed physical characteristics of the detector array, the maximum amount of information about the unknown object can be extracted from the measurement value; S3, executing an optimization algorithm: solving the optimization objective function by a numerical optimization algorithm, i.e., the maximum amount of information about the unknown object (145) to be imaged, to finally generate an optimal coded aperture pattern customized for the specific perovskite detector array (150).

6. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 5, characterized in that, The high-fidelity forward model in step S1 fully characterizes the whole-link physical process from X-ray emission, virtual focal spot generation, coded aperture modulation to detector capture, including virtual focal spot illumination field characteristics, coded aperture modulation law, photoelectric conversion characteristics of perovskite detector, noise characteristics and thermal load characteristics under high-brightness X-ray.

7. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 5, characterized in that, The optimization objective function expression in step S2 is: （1） in, This represents the optimal coded aperture pattern obtained from the final solution; It is a commonly used operator in mathematical optimization, meaning "in the encoding aperture parameter". Within the feasible range, find the one that maximizes the subsequent function value. ”; It refers to mutual information in information theory; Represents the true information of the unknown object to be imaged; Represents the coded measurements captured by the perovskite detector; Under the assumption that the imaging system is a linear model and the information of the object to be imaged and the noise are subject to Gaussian distribution, the mutual information The specific analytical expression of the mutual information is: （2） wherein: is a transmission matrix for the imaging system; is a noise covariance matrix; is an object prior covariance matrix; is an identity matrix; is a determinant operation on a matrix.

8. The perovskite imaging system based on coupling of high-brightness X-ray source according to claim 5, characterized in that, The numerical optimization algorithm in step S3 includes simulated annealing algorithm or genetic algorithm; and when the optimization algorithm in step S3 is used to design a mask, if it is detected that "multi-frame phase stepping method" will cause threshold overshoot due to lag, a "single exposure" type mask will be automatically selected. In step S3, the numerical optimization algorithm is solved under the constraint of the Nyquist theorem, the expression of the Nyquist theorem constraint is: ; wherein, is the mask unit size of the encoding aperture is the projection size on the detector array pixel after system optical magnification; is the pixel size of the detector array; In step S3, the numerical optimization algorithm is also subject to thermal stability constraints and coded aperture structure constraints; the thermal stability constraint is that the maximum temperature of the coded aperture pattern under the action of the virtual focal spot illumination field is less than or equal to the thermal stability limit temperature of the coded aperture material; and the coded aperture structure constraint is that the opening rate of the coded aperture pattern is greater than 0 and less than the engineering upper limit of the opening rate, and the unit size of the coded aperture pattern is between the processing precision limits of the unit size.

9. The perovskite imaging system based on coupling of high-brightness X-ray source according to any one of claims 1-8, characterized in that, The processing unit has stored therein customized pattern data of the coded aperture and a pre-constructed high-fidelity forward model; the reconstruction algorithm used by the processing unit includes maximum likelihood expectation maximization algorithm, algorithm based on compressed sensing or algorithm based on deep learning; and the spatial resolution of the imaging system is better than 3 lp / mm.

10. A perovskite imaging method based on coupling of high-brightness X-ray sources, characterized by, The application of the perovskite imaging system based on the high-brightness X-ray light source coupling according to any one of claims 1-9 comprises the following steps: S100, emitting X-ray radiation by a high-performance X-ray radiation source (110); S200, intercepting the X-ray radiation by using an optical divergence element (120) to generate a divergent light beam emitted from a virtual focal spot 125; S300, modulating the divergent light beam by a coded aperture (130) so that the modulated light beam irradiates a to-be-imaged object (145); S400, capturing the radiation after interaction with the to-be-imaged object (145) by using a detector array (150) based on perovskite material to generate coded measurement values; S500, solving the inverse problem of "inverting the real information of the object from the measurement values" by a processing unit (160) according to the preset coded aperture (130) pattern, system forward model and the coded measurement values, using a reconstruction algorithm, and outputting a high-resolution and high-fidelity image of the to-be-imaged object (145).

Images