Small coded aperture remote radioactive imaging system

Abstract

The invention particularly provides a small coded aperture remote radioactive imaging system. The system comprises a coding aperture camera module which is used for enabling incident gamma rays to form coding data and converting gamma ray photons into original electric signals; the FPGA data processing module is used for carrying out real-time parallel underlying processing on the original electric signal to obtain processed data; the digital multi-channel analysis module is used for performing energy analysis on the processed data to form energy spectrum data; the data storage module is used for storing collected original list data, intermediate data in processing and final reconstructed images and results; the data transmission module is used for data communication between modules in the system and between the system and peripherals; and the system data acquisition and processing module is used for controlling the execution sequence of each module, operating an imaging algorithm neural network and reconstructing coded data into a visual radioactive distribution image. And small coding aperture remote radioactivity rapid and accurate imaging is realized.

Description

Technical Field

[0001] This invention relates to the field of image data processing technology, specifically to the field of radioactive imaging technology, and specifically provides a small-sized coded aperture long-range radioactive imaging system. Background Technology

[0002] Mineral resource trade and processing occupy a pivotal position in the overall economy, making significant contributions to economic development, but also bringing environmental pollution risks. When supervising mineral resource utilization activities, environmental protection departments require the safe and efficient completion of component testing and radioactivity level monitoring for mineral raw materials, products, or waste residues with a certain storage area. Mines typically cover several square kilometers, and the targets to be tested (such as tailings piles and mineral raw materials) may exhibit diffuse and low-intensity characteristics. Existing portable instruments not only expose personnel to radiation risks but also, due to the sparse sampling points, easily miss local "hot spots."

[0003] Traditionally, mine radiation monitoring has relied mainly on portable monitoring instruments carried by personnel. While these instruments can basically meet the initial measurement needs of radiation dose, a series of problems have been exposed in practical applications: First, mines cover a large area, and manual monitoring is not only time-consuming, labor-intensive, and inefficient, but also difficult to achieve comprehensive coverage of the entire mine; second, long-term operation in a radiation environment, even with low-dose radiation, poses a potential threat to the health of monitoring personnel; and third, the sensitivity of portable monitoring instruments is limited, making it difficult to capture subtle changes in the radiation source, affecting the accuracy and completeness of the monitoring data.

[0004] While using small drones to carry detection equipment can solve the above problems, it is easily limited by the weight of the drones. Existing radiometric imaging systems are usually bulky, complex to operate, and suffer from low resolution and slow imaging speed in long-range imaging. Traditional coded aperture imaging devices, although highly accurate, typically use heavy lead / tungsten shielding and large-area detectors to ensure shielding effectiveness and detection efficiency, resulting in a total weight of over 10 kg, far exceeding the payload capacity of small and medium-sized drones. Simply reducing the size of the device (reducing the sensitive detection area and thinning the shielding layer) will lead to a sharp decrease in the number of gamma photons collected, resulting in a severe drop in the signal-to-noise ratio (SNR). Traditional correlation deconvolution algorithms will exhibit severe artifacts and low resolution problems when processing such sparse and noisy data.

[0005] Accordingly, there is a need in the art for a new small-sized coded aperture long-range radiometric imaging scheme to address the above problems. Summary of the Invention

[0006] To overcome the above-mentioned shortcomings, this invention is proposed to provide a small-sized coded aperture long-range radiographic imaging system that solves or at least partially solves the technical problems of existing radiographic imaging systems, such as large size, complex operation, low resolution and slow imaging speed in long-range imaging.

[0007] In a first aspect, the present invention provides a small-scale coded aperture long-range radiometric imaging system, the system being installed inside an airborne coded aperture camera, the system comprising a coded aperture camera module, an FPGA data processing module, a digital multichannel analysis module, a data storage module, a data transmission module, and a system data acquisition and processing module. The coded aperture camera module is used to form coded data from incident gamma rays and to convert gamma ray photons into raw electrical signals. The FPGA data processing module is used to perform real-time, parallel low-level processing on the original electrical signals to obtain the processed data. The digital multichannel analysis module is used to perform energy analysis on the processed data to generate energy spectrum data; The data storage module is used to store the collected raw list data, intermediate data during processing, and the final reconstructed image and result; The data transmission module connects to each module and is used for data communication between modules within the system and between the system and peripherals. The system data acquisition and processing module controls the execution order of each module and runs the imaging algorithm neural network to reconstruct the encoded data into a visualized image of the radioactive distribution.

[0008] In one technical solution of the aforementioned small-scale coded aperture long-range radiometric imaging system, the coded aperture camera module includes a radiation detector, and the coded aperture camera module: The incident gamma rays are spatially modulated by an coded aperture plate to form coded data, wherein the coded data is a projection pattern carrying directional information. The radiation detector converts gamma-ray photons into raw electrical signals.

[0009] In one technical solution of the aforementioned small-aperture remote radiographic imaging system, the FPGA data processing module: All raw electrical signals output by the radiation detector are processed in real time and in parallel at the underlying level. The underlying processing includes at least pulse shaping, amplitude extraction, timestamp marking, preliminary filtering and discrimination to obtain processed data.

[0010] In one technical solution of the aforementioned small-aperture long-range radiometric imaging system, the digital multichannel analysis module: Energy analysis is performed on the pulse signals in the processed data output by the FPGA data processing module. The energy analysis includes classifying and counting the pulses according to their amplitude to form energy spectrum data.

[0011] In one technical solution of the aforementioned small-code aperture remote radiographic imaging system, the raw list data stored in the data storage module includes at least: time, energy, and location.

[0012] In one technical solution of the aforementioned small-aperture long-range radioactive imaging system, running the imaging algorithm neural network includes the following steps: Retrieve raw list data, intermediate data under processing, and encoded data stored in the data storage module; Based on the original list data, intermediate data in processing, and encoded data, a visualized image of the radioactive distribution is obtained by mapping and reconstructing the data using a trained imaging algorithm neural network.

[0013] In one technical solution of the aforementioned small-aperture long-range radioactive imaging system, training the imaging algorithm neural network includes the following steps: Obtain multiple sets of design parameters, wherein the design parameter sets include at least the aperture ratio, aperture size, aperture arrangement, plate thickness, and array order of the coding plate; Imaging process simulation is performed using Monte Carlo simulation based on multiple sets of design parameters to obtain multiple sets of simulation datasets. The simulation datasets include at least source term distribution data and encoded image data. Acquire a variety of artificial intelligence models, including at least ResNet, GAN, CNN, BPNN, LSTM, and GRU; Different types of artificial intelligence models were trained and tested based on multiple sets of simulated datasets to obtain various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model. To obtain a trained imaging algorithm neural network.

[0014] In one technical solution of the aforementioned small-aperture long-range radiographic imaging system, the training and testing of different types of artificial intelligence models based on multiple sets of simulated datasets, resulting in various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model, includes: Multiple sets of simulated datasets are divided into multiple training sets and multiple test sets according to a preset ratio; Multiple training sets are substituted into different types of artificial intelligence models to obtain the reconstructed image data corresponding to each training set. Substitute the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data; Based on multiple sets of analysis results, a well-trained artificial intelligence model can be selectively obtained, or the artificial intelligence model can be optimized and iteratively trained again on the optimized artificial intelligence model. This allows us to obtain a variety of trained artificial intelligence models and the range of design parameter sets corresponding to each model.

[0015] In one technical solution of the aforementioned small-aperture long-range radiographic imaging system, the step of substituting the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data includes: The reconstructed image data is denoised by Gaussian filtering using a filtering algorithm to obtain the filtered reconstructed image data. Edge detection algorithms are used to perform edge recognition on the filtered reconstructed image data to obtain reconstructed image data after edge recognition. The reconstructed image data after edge recognition is transformed by a transformation algorithm to obtain the transformed reconstructed image data. The missing data of the transformed reconstructed image data is eliminated by interpolation algorithm to obtain complete reconstructed image data; This is to obtain the analysis results corresponding to each set of reconstructed image data.

[0016] In one technical solution of the aforementioned small-aperture long-range radiographic imaging system, the basis for optimizing the artificial intelligence model includes: Based on the analysis results corresponding to each set of reconstructed image data, the parameters of imaging field of view, spatial angular resolution, peak signal-to-noise ratio, mean square error, and spatial positioning accuracy corresponding to each set of reconstructed image data are obtained. The parameter values ​​of different artificial intelligence models are compared to identify the key factors that significantly affect imaging accuracy, and the key factors are optimized.

[0017] The above-described technical solutions of the present invention have at least one or more of the following beneficial effects: This invention achieves a miniaturized and lightweight design for an airborne coded aperture camera through compact and lightweight construction and high-efficiency modular integration. Simultaneously, the introduction of artificial intelligence algorithms into coded aperture imaging technology improves the imaging speed and measurement accuracy of the camera, enabling rapid and accurate remote radioactive imaging with a small coded aperture. This system boasts advantages such as small size, light weight, high imaging accuracy, and ease of operation, making it particularly suitable for remote radioactive imaging detection. It improves imaging resolution, meeting high-precision imaging requirements. The system's overall small size and light weight facilitate portability and deployment, making it suitable for large-scale rapid radioactive detection in various complex environments. While ensuring personnel safety, it also enables remote operation and data transmission, facilitating remote monitoring and detection for users. Attached Figure Description

[0018] The disclosure of this invention will become more readily understood with reference to the accompanying drawings. It will be readily understood by those skilled in the art that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. Furthermore, similar numbers in the drawings are used to denote similar components, wherein: Figure 1 This is a schematic diagram of the imaging process of a small coded aperture long-range radiometric imaging system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the main structure of a small coded aperture long-range radiographic imaging system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the system casing of a small coded aperture remote radiographic imaging system according to an embodiment of the present invention. Detailed Implementation

[0019] Some embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0020] In the description of this invention, "module" and "processor" can include hardware, software, or a combination of both. A module can include hardware circuitry, various suitable sensors, communication ports, memory, and may also include software components, such as program code, or a combination of software and hardware. A processor can be a central processing unit, microprocessor, image processor, digital signal processor, or any other suitable processor. The processor has data and / or signal processing capabilities. The processor can be implemented in software, in hardware, or a combination of both. Non-transitory computer-readable storage media includes any suitable medium capable of storing program code, such as magnetic disks, hard disks, optical disks, flash memory, read-only memory, random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as only A, only B, or A and B. The terms "at least one A or B" or "at least one of A and B" have a similar meaning to "A and / or B" and can include only A, only B, or A and B. The singular terms "a" or "this" can also include plural forms.

[0021] The directional terms used in this article, such as "front," "front side," "front part," "rear," "rear side," and "rear part," are all based on the front-rear direction of the vehicle after the component is installed. The terms "longitudinal," "longitudinal direction," and "longitudinal section" mentioned in this article are based on the front-rear direction after the component is installed in the vehicle, while "transverse," "lateral," and "cross section" indicate the direction perpendicular to the longitudinal direction.

[0022] See appendix Figures 1 to 2 , Figure 2 This is a schematic diagram of the main structure of a small-code aperture long-range radiographic imaging system according to an embodiment of the present invention. Figures 1 to 2 As shown, the small-scale coded aperture remote radiometric imaging system in this embodiment of the invention is installed inside an airborne coded aperture camera. The system includes a coded aperture camera module, an FPGA data processing module, a digital multichannel analysis module, a data storage module, a data transmission module, and a system data acquisition and processing module. The coded aperture camera module is used to form coded data from incident gamma rays and to convert gamma ray photons into raw electrical signals. The FPGA data processing module is used to perform real-time, parallel low-level processing on the original electrical signals to obtain the processed data. The digital multichannel analysis module is used to perform energy analysis on the processed data to generate energy spectrum data; The data storage module is used to store the collected raw list data, intermediate data during processing, and the final reconstructed image and result; The data transmission module connects to each module and is used for data communication between modules within the system and between the system and peripherals. The system data acquisition and processing module controls the execution order of each module and runs the imaging algorithm neural network to reconstruct the encoded data into a visualized image of the radioactive distribution.

[0023] Specifically, the coded aperture camera module includes a radiation detector. The coded aperture camera module: The incident gamma rays are spatially modulated by an coded aperture plate to form coded data, wherein the coded data is a projection pattern carrying directional information. The radiation detector converts gamma-ray photons into raw electrical signals. The original electrical signal is then sent to the FPGA data processing module via the data transmission module. The encoded data is then sent to the system data acquisition and processing module via the data transmission module.

[0024] Specifically, the coded aperture plate is made of a material with low transmittance and high radiation absorption rate, such as heavy metal or polymer shielding material. The advantage of using such material is that it can shield / block rays as much as possible, preventing rays from penetrating the material. That is, if rays penetrate the material, they will be detected by the detector, thus causing errors. Tungsten, brass, etc. can be used. The aperture distribution on the coded aperture plate follows specific coding rules, such as random coding, pseudo-random coding, etc., to maximize information content and imaging quality.

[0025] Specifically, the detector employs a high-sensitivity, low-noise scintillation crystal or semiconductor array detector, such as a silicon PIN detector, germanium detector, or zinc-chromium telluride detector; and the pixel size and number of the detector array are set according to the imaging area and resolution requirements.

[0026] Specifically, the FPGA data processing module: All raw electrical signals output by the radiation detector are processed in real time and in parallel at the low level. The low-level processing includes at least pulse shaping, amplitude extraction, timestamp marking, preliminary filtering and discrimination to obtain processed data. The processed data is then sent to the digital multichannel analysis module.

[0027] Specifically, the digital multichannel analysis module: Energy analysis is performed on the pulse signals in the processed data output by the FPGA data processing module. The energy analysis includes classifying and counting the pulses according to their amplitude to form energy spectrum data.

[0028] Specifically, the raw list data stored in the data storage module includes at least: time, energy, and location.

[0029] Specifically, such as Figure 3As shown, the system's outer shell features a low-drag streamlined design with a smooth surface transition, reducing flight drag and airflow noise to avoid affecting the drone's endurance and flight stability. It is also constructed from lightweight, high-strength carbon fiber composite material, achieving extreme lightweighting while maintaining structural strength to meet the drone's payload requirements. The internal design incorporates modular mounting slots and shock-absorbing structures to ensure precise alignment and stable operation of all functional modules, such as the encoder plate, detectors, and circuit boards, in flight vibration environments. Furthermore, multiple symmetrically distributed slender cylindrical mounting supports are located on the top of the shell, facilitating system mounting on the drone while maintaining high axial and radial stiffness.

[0030] In addition, the shell has reserved standard drone mounting interfaces and quick-release structures to support flexible mounting and quick disassembly of various drone models, facilitating field deployment and equipment maintenance.

[0031] Specifically, the system adopts a lightweight and high-strength frame structure. By selecting lightweight, high-strength, and corrosion-resistant carbon fiber frame and alloy structural materials, and combining the design principle based on structural lightweighting, the lightweight design of the airborne coded aperture camera is realized, reducing the overall weight of the equipment and improving its durability and system battery life.

[0032] Specifically, the wireless communication module includes an embedded processor, a wireless communication module, and a human-machine interface: The processor is responsible for data processing, decoding algorithm execution, and receiving and executing remote control commands; The wireless communication module supports Wi-Fi, Bluetooth, or 4G / 5G networks to enable remote data transmission and real-time monitoring. Human-machine interfaces can be touchscreens, keyboards, or remote computer terminals, allowing users to perform real-time monitoring and data analysis from anywhere.

[0033] Specifically, running the imaging algorithm neural network includes the following steps: Retrieve raw list data, intermediate data under processing, and encoded data stored in the data storage module; Based on the original list data, intermediate data in processing, and encoded data, a visualized image of the radioactive distribution is obtained by mapping and reconstructing the data using a trained imaging algorithm neural network.

[0034] In the above embodiments, the imaging algorithm neural network is specifically designed to process projection data with known degradation characteristics obtained from a specific miniaturized coded aperture system. Its input is a low signal-to-noise ratio, low-resolution aliased projection image, and its output is a clear, high-resolution image of the radiation source distribution. This enables the neural network to perform real-time inference on an airborne computing unit while miniaturizing the network and ensuring image accuracy while reducing inference time.

[0035] Specifically, training the imaging algorithm neural network includes the following steps: Obtain multiple sets of design parameters, wherein the design parameter sets include at least the aperture ratio, aperture size, aperture arrangement, plate thickness, and array order of the coding plate; Imaging process simulation is performed using Monte Carlo simulation based on multiple sets of design parameters to obtain multiple sets of simulation datasets. The simulation datasets include at least source term distribution data and encoded image data. Acquire a variety of artificial intelligence models, including at least ResNet, GAN, CNN, BPNN, LSTM, and GRU; Different types of artificial intelligence models were trained and tested based on multiple sets of simulated datasets to obtain various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model. To obtain a trained imaging algorithm neural network.

[0036] Specifically, the training and testing of different types of artificial intelligence models based on multiple sets of simulated datasets to obtain various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model includes: Multiple sets of simulated datasets are divided into multiple training sets and multiple test sets according to a preset ratio; Multiple training sets are substituted into different types of artificial intelligence models to obtain the reconstructed image data corresponding to each training set. Substitute the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data; Based on multiple sets of analysis results, a well-trained artificial intelligence model can be selectively obtained, or the artificial intelligence model can be optimized and iteratively trained again on the optimized artificial intelligence model. This allows us to obtain a variety of trained artificial intelligence models and the range of design parameter sets corresponding to each model.

[0037] Specifically, the step of substituting the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data includes: The reconstructed image data is denoised by Gaussian filtering using a filtering algorithm to obtain the filtered reconstructed image data. Edge detection algorithms are used to perform edge recognition on the filtered reconstructed image data to obtain reconstructed image data after edge recognition. The reconstructed image data after edge recognition is transformed by a transformation algorithm to obtain the transformed reconstructed image data. The missing data of the transformed reconstructed image data is eliminated by interpolation algorithm to obtain complete reconstructed image data; This is to obtain the analysis results corresponding to each set of reconstructed image data.

[0038] Specifically, in this embodiment, the execution of each algorithm in the above imaging analysis model adopts a conventional execution scheme. The execution scheme settings of each algorithm in the imaging analysis model are only illustrative examples. Those skilled in the art can set them according to actual usage needs, as long as the analysis results corresponding to each set of reconstructed image data can be obtained through the imaging analysis model. Further details are omitted here.

[0039] Specifically, the basis for optimizing the artificial intelligence model includes: Based on the analysis results corresponding to each set of reconstructed image data, the parameters of imaging field of view, spatial angular resolution, peak signal-to-noise ratio, mean square error, and spatial positioning accuracy corresponding to each set of reconstructed image data are obtained. The parameter values ​​of different artificial intelligence models are compared to identify the key factors that significantly affect imaging accuracy, and the key factors are optimized.

[0040] Specifically, in some embodiments, the present invention employs the Monte Carlo method to simulate the γ-ray coded aperture imaging process under different system parameters. By arbitrarily adjusting design parameters that may significantly affect positioning accuracy in the Monte Carlo model, such as the aperture ratio, aperture size, aperture arrangement, plate thickness, and array order of the coded aperture plate, a large amount of simulated imaging data is obtained, including source term distribution data and coded image data. This simulated dataset is then used to train and test artificial intelligence models, such as ResNet, GAN, CNN, BPNN, LSTM, and GRU network models. Complete training, debugging, and verification of the artificial intelligence model require... To ensure sufficient sample quantity, sample diversity, and effective sample features, and combined with image processing and interpolation algorithms such as filtering, edge detection, and transformation algorithms, a system imaging effect evaluation method is established. By analyzing the changes in parameters such as imaging field of view, spatial angular resolution, peak signal-to-noise ratio, mean square error, and spatial positioning accuracy as design parameters are adjusted, different image imaging results are analyzed. From the variation patterns of the analysis results, key factors that significantly affect imaging accuracy are identified, and their degree of influence is explored and quantitatively analyzed. The coded aperture imaging mechanism based on artificial intelligence algorithms is revealed, effectively improving the imaging speed and accuracy of miniaturized gamma cameras.

[0041] Specifically, the construction and training of artificial intelligence models includes: The network architecture of this system adopts an imaging reconstruction architecture based on an improved artificial neural network. The core network structure includes: Input layer: Receives a pre-processed two-dimensional coded projection image acquired by the coded aperture camera module. The input size is N×N pixels and the number of channels is 1 (grayscale image). Feature extraction module: consists of 4 convolutional blocks, each of which includes: convolutional layer: kernel size 3×3, stride 1, padding 1; batch normalization layer: accelerates convergence and prevents overfitting; ReLU activation function: introduces non-linearity; max pooling layer: pooling window size 2×2, stride 2, gradually reducing feature map resolution; Attention mechanism module: Channel attention modules are inserted after the third and fourth convolutional blocks to enhance the response to key energy channels and improve reconstruction accuracy; Decoding and reconstruction module: consists of 3 upsampling blocks, each upsampling block includes: transposed convolutional layer: convolutional kernel size is 3×3, stride is 2; skip connection: concatenates the feature maps of the corresponding encoder layer to preserve detailed information; ReLU activation function; Output layer: A convolutional layer with a kernel size of 1×1 and 1 output channel is used. The activation function is Sigmoid to generate an N×N pixel reconstructed radioactive distribution image.

[0042] The inputs and outputs of this system include: The input data consists of simulated or measured coded image data, which is normalized to the [0,1] interval and used as network input. The output data is a distribution image of the radioactive source terms corresponding to the input, which serves as the supervision signal for the network output. The training objective is to minimize the error between the output image and the true distribution image.

[0043] The training process and key parameters of this system include: Dataset construction: Using Monte Carlo simulation tools (such as Geant4 or MCNP), imaging data pairs were generated under different coded perforation plate design parameters (including at least aperture ratio, thickness, array order, etc.), namely source term distribution maps and coded images. A total of 100,000 sets of samples were generated and divided into training set, validation set and test set in an 8:1:1 ratio. Training epochs: 100 epochs in total, batch size set to 32; Early stopping mechanism: Training is stopped when the validation set loss does not decrease for 10 consecutive epochs to prevent overfitting.

[0044] The model evaluation and optimization of this system are based on the following: The imaging field of view (FOV), spatial angular resolution, peak signal-to-noise ratio (PSNR), mean square error (MSE), structural similarity (SSIM), and spatial positioning accuracy of the model were evaluated on the test set. Compare the performance of different network structures (such as ResNet, GAN, LSTM, etc.) on the above metrics and select the optimal model; Based on the analysis results, key factors such as network depth, attention module location, and loss function weights were further adjusted to iteratively optimize the network structure.

[0045] Specifically, the system's working process is as follows: When the radiation emitted by the radiation source, specifically gamma rays in this case, passes through the coded aperture plate, each aperture on the plate generates an image of the object on the detector. All these projections overlap, which is the encoding process. All the projection signals are received, stored, processed, and output by the signal acquisition system to generate a two-dimensional coded image. This image cannot capture the original source information; therefore, it is necessary to infer the original spatial distribution image of the radiation source based on the overlap information of the aperture positions—this is the decoding process. Specifically, the decoding process includes the following steps: 1) X-ray encoding: code aperture, while shielding other directions. This is a key step in spatial modulation to achieve imaging: the encoding aperture plate is made of a high atomic number material, such as tungsten or lead; when the X-rays emitted by the radiation source irradiate the encoding aperture plate, the encoding aperture plate acts as a "template" or "modulator", allowing some X-rays to pass through the specific encoded direction of the X-rays, so that the source point information in each direction is encoded into a projection with the pattern features of the encoding aperture plate; 2) X-ray detection: The encoded X-ray beam, carrying modulated spatial information, illuminates a large-area radiation detector. The detector in the encoded aperture camera module converts the invisible X-ray photons into measurable electrical signals. The detector may include a scintillator detector (including one of sodium iodide, lanthanum bromide coupled photomultiplier tube, or silicon photomultiplier tube) and a semiconductor detector (including one of zinc cadmium telluride or high-purity germanium). Each incident photon deposits energy in the detector, generating an optical signal or an electron-hole pair, which is then converted into an electrical pulse signal with an amplitude proportional to the photon energy. 3) Signal Processing: The raw electrical pulse signal output by the coded aperture camera module is very weak and accompanied by noise. Therefore, it needs to undergo a series of electronic processing steps before it can be used for imaging. This process first involves preliminary amplification and impedance matching by a preamplifier. Then, the pulse is further amplified and the waveform is optimized by the main amplifier / shaping circuit in the FPGA data processing module to improve the signal-to-noise ratio. Next, the analog-to-digital converter in the FPGA data processing module accurately converts the amplitude, i.e., energy and time information of the analog pulse into digital signals. These digital signals are received by a multichannel analyzer or a dedicated digital acquisition card, and are identified and classified according to energy and time. Finally, a list-mode data containing the energy and timestamp of each event is formed, i.e., the processed data. 4) Decoding and Imaging: The list data received by the digital multichannel analysis module, i.e. the processed data, is first converted into two-dimensional projection data (i.e., encoded image) corresponding to the geometry of the encoded aperture plate; then, the system data acquisition and processing module runs the decoding and reconstruction algorithm: The deep learning algorithm based on artificial neural networks used in this invention is trained with a large amount of simulation and experimental data. It can directly and quickly and accurately map the final image from the encoded image or intermediate features. It not only improves the imaging speed and enables near real-time imaging, but also effectively suppresses noise, reduces artifacts, and improves the imaging quality under low count rate or complex background conditions. 5) Remote monitoring: The imaging results are transmitted to a remote computer terminal or mobile device via a wireless communication module, allowing users to perform real-time monitoring and data analysis from anywhere.

[0046] Based on the aforementioned small-sized coded aperture remote radioactivity imaging system, this invention achieves a miniaturized and lightweight design of the airborne coded aperture camera through compact and lightweight design and high-efficiency modular integration. Simultaneously, the introduction of artificial intelligence algorithms into coded aperture imaging technology improves the imaging speed and measurement accuracy of the coded aperture camera, enabling rapid and accurate remote radioactivity imaging with a small coded aperture. This system boasts advantages such as small size, light weight, high imaging accuracy, and ease of operation, making it particularly suitable for remote radioactivity imaging detection. It improves imaging resolution, meeting high-precision imaging requirements. The system's overall small size and light weight facilitate portability and deployment, making it suitable for large-scale rapid radioactivity detection in various complex environments. While ensuring personnel safety, it enables remote operation and data transmission, facilitating remote monitoring and detection for users.

[0047] It should be noted that although the steps in the above embodiments are described in a specific order, those skilled in the art will understand that in order to achieve the effects of the present invention, different steps do not necessarily have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders, and these variations are all within the scope of protection of the present invention.

[0048] Those skilled in the art will understand that all or part of the processes in the method of the above embodiment of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium can include any entity or device capable of carrying the computer program code, a medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the content included in the computer-readable storage medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable storage medium does not include electrical carrier signals and telecommunication signals.

[0049] Furthermore, the present invention also provides a control device. In one embodiment of the control device according to the present invention, the control device includes a processor and a storage device. The storage device can be configured to store a program for executing the small coded aperture remote radiographic imaging system of the above-described method embodiments. The processor can be configured to execute the program in the storage device, which includes, but is not limited to, the program for executing the small coded aperture remote radiographic imaging system of the above-described method embodiments. For ease of explanation, only the parts related to the embodiments of the present invention are shown; for specific technical details not disclosed, please refer to the method section of the embodiments of the present invention. This control device can be a control device device comprising various electronic devices.

[0050] Furthermore, the present invention also provides a computer-readable storage medium. In one embodiment of the computer-readable storage medium according to the present invention, the computer-readable storage medium can be configured to store a program for executing a small coded aperture remote radiographic imaging system according to the above-described method embodiments. This program can be loaded and run by a processor to implement the operation process of the small coded aperture remote radiographic imaging system. For ease of explanation, only the parts related to the embodiments of the present invention are shown; for specific technical details not disclosed, please refer to the method section of the embodiments of the present invention. The computer-readable storage medium can be a storage device comprising various electronic devices. Optionally, in the embodiments of the present invention, the computer-readable storage medium is a non-transitory computer-readable storage medium.

[0051] Furthermore, it should be understood that since the various modules are only provided to illustrate the functional units of the device of the present invention, the physical devices corresponding to these modules may be the processor itself, or a part of the processor's software, hardware, or a combination of software and hardware. Therefore, the number of modules shown in the figures is merely illustrative.

[0052] Those skilled in the art will understand that the various modules in the device can be adaptively split or combined. Such splitting or combining of specific modules will not cause the technical solution to deviate from the principles of the present invention; therefore, the technical solutions after splitting or combining will fall within the protection scope of the present invention.

[0053] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A small-sized coded aperture long-range radiometric imaging system, characterized in that, The system is installed inside an airborne coded aperture camera and includes a coded aperture camera module, an FPGA data processing module, a digital multichannel analysis module, a data storage module, a data transmission module, and a system data acquisition and processing module. The coded aperture camera module is used to form coded data from incident gamma rays and to convert gamma ray photons into raw electrical signals. The FPGA data processing module is used to perform real-time, parallel low-level processing on the original electrical signals to obtain the processed data. The digital multichannel analysis module is used to perform energy analysis on the processed data to generate energy spectrum data; The data storage module is used to store the collected raw list data, intermediate data during processing, and the final reconstructed image and result; The data transmission module connects to each module and is used for data communication between modules within the system and between the system and peripherals. The system data acquisition and processing module controls the execution order of each module and runs the imaging algorithm neural network to reconstruct the encoded data into a visualized image of the radioactive distribution.

2. The small-sized coded aperture long-range radiographic imaging system according to claim 1, characterized in that, The coded aperture camera module includes a radiation detector. The coded aperture camera module: The incident gamma rays are spatially modulated by an coded aperture plate to form coded data, wherein the coded data is a projection pattern carrying directional information. The radiation detector converts gamma-ray photons into raw electrical signals.

3. The small-sized coded aperture long-range radiometric imaging system according to claim 2, characterized in that, The FPGA data processing module: All raw electrical signals output by the radiation detector are processed in real time and in parallel at the underlying level. The underlying processing includes at least pulse shaping, amplitude extraction, timestamp marking, preliminary filtering and discrimination to obtain processed data.

4. The small-sized coded aperture long-range radiometric imaging system according to claim 3, characterized in that, The digital multichannel analysis module: Energy analysis is performed on the pulse signals in the processed data output by the FPGA data processing module. The energy analysis includes classifying and counting the pulses according to their amplitude to form energy spectrum data.

5. The small-sized coded aperture long-range radiographic imaging system according to claim 4, characterized in that, The raw list data stored in the data storage module includes at least: time, energy, and location.

6. The small-sized coded aperture long-range radiometric imaging system according to claim 5, characterized in that, Running the imaging algorithm neural network includes the following steps: Retrieve raw list data, intermediate data under processing, and encoded data stored in the data storage module; Based on the original list data, intermediate data in processing, and encoded data, a visualized image of the radioactive distribution is obtained by mapping and reconstructing the data using a trained imaging algorithm neural network.

7. The small coded aperture long-range radiometric imaging system according to claim 6, characterized in that, Training the imaging algorithm neural network includes the following steps: Obtain multiple sets of design parameters, wherein the design parameter sets include at least the aperture ratio, aperture size, aperture arrangement, plate thickness, and array order of the coding plate; Imaging process simulation is performed using Monte Carlo simulation based on multiple sets of design parameters to obtain multiple sets of simulation datasets. The simulation datasets include at least source term distribution data and encoded image data. Acquire a variety of artificial intelligence models, including at least ResNet, GAN, CNN, BPNN, LSTM, and GRU; Different types of artificial intelligence models were trained and tested based on multiple sets of simulated datasets to obtain various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model. To obtain a trained imaging algorithm neural network.

8. The small coded aperture long-range radiometric imaging system according to claim 7, characterized in that, The process of training and testing different types of artificial intelligence models based on multiple sets of simulated datasets to obtain various trained artificial intelligence models and the range of design parameter sets corresponding to each artificial intelligence model includes: Multiple sets of simulated datasets are divided into multiple training sets and multiple test sets according to a preset ratio; Multiple training sets are substituted into different types of artificial intelligence models to obtain the reconstructed image data corresponding to each training set. Substitute the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data; Based on multiple sets of analysis results, a well-trained artificial intelligence model can be selectively obtained, or the artificial intelligence model can be optimized and iteratively trained again on the optimized artificial intelligence model. This allows us to obtain a variety of trained artificial intelligence models and the range of design parameter sets corresponding to each model.

9. The small-sized coded aperture long-range radiometric imaging system according to claim 8, characterized in that, The step of substituting the reconstructed image data corresponding to each training set into the imaging analysis model to obtain the analysis results corresponding to each set of reconstructed image data includes: The reconstructed image data is denoised by Gaussian filtering using a filtering algorithm to obtain the filtered reconstructed image data. Edge detection algorithms are used to perform edge recognition on the filtered reconstructed image data to obtain reconstructed image data after edge recognition. The reconstructed image data after edge recognition is transformed by a transformation algorithm to obtain the transformed reconstructed image data. The missing data of the transformed reconstructed image data is eliminated by interpolation algorithm to obtain complete reconstructed image data; This is to obtain the analysis results corresponding to each set of reconstructed image data.

10. The small-sized coded aperture long-range radiometric imaging system according to claim 8, characterized in that, The basis for optimizing the artificial intelligence model includes: Based on the analysis results corresponding to each set of reconstructed image data, the parameters of imaging field of view, spatial angular resolution, peak signal-to-noise ratio, mean square error, and spatial positioning accuracy corresponding to each set of reconstructed image data are obtained. The parameter values ​​of different artificial intelligence models are compared to identify the key factors that significantly affect imaging accuracy, and the key factors are optimized.

Images