Image super-resolution reconstruction technology is currently widely used in medical imaging,video surveillance,and industrial quality inspection.It not only enhances image quality but also improves details and visual...Image super-resolution reconstruction technology is currently widely used in medical imaging,video surveillance,and industrial quality inspection.It not only enhances image quality but also improves details and visual perception,significantly increasing the utility of low-resolution images.In this study,an improved image superresolution reconstruction model based on Generative Adversarial Networks(SRGAN)was proposed.This model introduced a channel and spatial attention mechanism(CSAB)in the generator,allowing it to effectively leverage the information from the input image to enhance feature representations and capture important details.The discriminator was designed with an improved PatchGAN architecture,which more accurately captured local details and texture information of the image.With these enhanced generator and discriminator architectures and an optimized loss function design,this method demonstrated superior performance in image quality assessment metrics.Experimental results showed that this model outperforms traditional methods,presenting more detailed and realistic image details in the visual effects.展开更多
In agricultural production,crop images are commonly used for the classification and identification of various crops.However,several challenges arise,including low image clarity,elevated noise levels,low accuracy,and p...In agricultural production,crop images are commonly used for the classification and identification of various crops.However,several challenges arise,including low image clarity,elevated noise levels,low accuracy,and poor robustness of existing classification models.To address these issues,this research proposes an innovative crop image classification model named Lap-FEHRNet,which integrates a Laplacian Pyramid Super Resolution Network(LapSRN)with a feature enhancement high-resolution network based on attention mechanisms(FEHRNet).To mitigate noise interference,this research incorporates the LapSRN network,which utilizes a Laplacian pyramid structure to extract multi-level feature details from low-resolution images through a systematic layer-by-layer amplification and pixel detail superposition process.This gradual reconstruction enhances the high-frequency information of the image,enabling super-resolution reconstruction of low-quality images.To obtain a broader range of comprehensive and diverse features,this research employs the FEHRNetmodel for both deep and shallow feature extraction.This approach results in features that encapsulate multi-scale information and integrate both deep and shallow insights.To effectively fuse these complementary features,this research introduces an attention mechanism during the feature enhancement stage.This mechanism highlights important regions within the image,assigning greater weights to salient features and resulting in a more comprehensive and effective image feature representation.Consequently,the accuracy of image classification is significantly improved.Experimental results demonstrate that the Lap-FEHRNetmodel achieves impressive classification accuracies of 98.8%on the crop classification dataset and 98.57%on the rice leaf disease dataset,underscoring the model’s outstanding accuracy,robustness,and generalization capability.展开更多
As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capaci...As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capacity.However,existing VQ-VAEs often perform quantization in the spatial domain,ignoring global structural information and potentially suffering from codebook collapse and information coupling issues.This paper proposes a frequency quantized variational autoencoder(FQ-VAE)to address these issues.The proposed method transforms image features into linear combinations in the frequency domain using a 2D fast Fourier transform(2D-FFT)and performs adaptive quantization on these frequency components to preserve image’s global relationships.The codebook is dynamically optimized to avoid collapse and information coupling issue by considering the usage frequency and dependency of code vectors.Furthermore,we introduce a post-processing module based on graph convolutional networks to further improve reconstruction quality.Experimental results on four public datasets demonstrate that the proposed method outperforms state-of-the-art approaches in terms of Structural Similarity Index(SSIM),Learned Perceptual Image Patch Similarity(LPIPS),and Reconstruction Fréchet Inception Distance(rFID).In the experiments on the CIFAR-10 dataset,compared to the baselinemethod VQ-VAE,the proposedmethod improves the abovemetrics by 4.9%,36.4%,and 52.8%,respectively.展开更多
Deep learning(DL)-based image reconstruction methods have garnered increasing interest in the last few years.Numerous studies demonstrate that DL-based reconstruction methods function admirably in optical tomographic ...Deep learning(DL)-based image reconstruction methods have garnered increasing interest in the last few years.Numerous studies demonstrate that DL-based reconstruction methods function admirably in optical tomographic imaging techniques,such as bioluminescence tomography(BLT).Nevertheless,nearly every existing DL-based method utilizes an explicit neural representation for the reconstruction problem,which either consumes much memory space or requires various complicated computations.In this paper,we present a neural field(NF)-based image reconstruction scheme for BLT that uses an implicit neural representation.The proposed NFbased method establishes a transformation between the coordinate of an arbitrary spatial point and the source value of the point with a relatively light-weight multilayer perceptron,which has remarkable computational efficiency.Another simple neural network composed of two fully connected layers and a 1D convolutional layer is used to generate the neural features.Results of simulations and experiments show that the proposed NF-based method has similar performance to the photon density complement network and the two-stage network,while consuming fewer floating point operations with fewer model parameters.展开更多
Person Image Synthesis has been widely used in fashion with extensive application scenarios.The point of this task is how to synthesise person image from a single source image under arbitrary poses.Prior methods gener...Person Image Synthesis has been widely used in fashion with extensive application scenarios.The point of this task is how to synthesise person image from a single source image under arbitrary poses.Prior methods generate the person image with target pose well;however,they fail to preserve the fine style details of the source image.To address this problem,a robust style injection(RSI)model is proposed,which is a coarse-to-fine framework to synthesise target the person image.RSI develops a simple and efficient cross-attention based module to fuse the features of both source semantic styles and target pose for achieving the coarse aligned features.The adaptive instance normalisation is employed to enhance the aligned features in conjunction with source semantic styles.Subsequently,source semantic styles are further injected into the positional normalisation scheme to avoid the fine style details erosion caused by massive convolution.In training losses,optimal transport theory in the form of energy distance is introduced to constrain data distribution to refine the texture style details.Additionally,the authors’model is capable of editing the shape and texture of garments to the target style separately.The experiments demonstrate that the authors’RSI achieves better performance over the state-of-art methods.展开更多
Accurate segmentation of camouflage objects in aerial imagery is vital for improving the efficiency of UAV-based reconnaissance and rescue missions.However,camouflage object segmentation is increasingly challenging du...Accurate segmentation of camouflage objects in aerial imagery is vital for improving the efficiency of UAV-based reconnaissance and rescue missions.However,camouflage object segmentation is increasingly challenging due to advances in both camouflage materials and biological mimicry.Although multispectral-RGB based technology shows promise,conventional dual-aperture multispectral-RGB imaging systems are constrained by imprecise and time-consuming registration and fusion across different modalities,limiting their performance.Here,we propose the Reconstructed Multispectral-RGB Fusion Network(RMRF-Net),which reconstructs RGB images into multispectral ones,enabling efficient multimodal segmentation using only an RGB camera.Specifically,RMRF-Net employs a divergentsimilarity feature correction strategy to minimize reconstruction errors and includes an efficient boundary-aware decoder to enhance object contours.Notably,we establish the first real-world aerial multispectral-RGB semantic segmentation of camouflage objects dataset,including 11 object categories.Experimental results demonstrate that RMRF-Net outperforms existing methods,achieving 17.38 FPS on the NVIDIA Jetson AGX Orin,with only a 0.96%drop in mIoU compared to the RTX 3090,showing its practical applicability in multimodal remote sensing.展开更多
A new method to accelerate the convergent rate of the space-alternatinggeneralized expectation-maximization (SAGE) algorithm is proposed. The new rescaled block-iterativeSAGE (RBI-SAGE) algorithm combines the RBI algo...A new method to accelerate the convergent rate of the space-alternatinggeneralized expectation-maximization (SAGE) algorithm is proposed. The new rescaled block-iterativeSAGE (RBI-SAGE) algorithm combines the RBI algorithm with the SAGE algorithm for PET imagereconstruction. In the new approach, the projection data is partitioned into disjoint blocks; eachiteration step involves only one of these blocks. SAGE updates the parameters sequentially in eachblock. In experiments, the RBI-SAGE algorithm and classical SAGE algorithm are compared in theapplication on positron emission tomography (PET) image reconstruction. Simulation results show thatRBI-SAGE has better performance than SAGE in both convergence and image quality.展开更多
In high-resolution cone-beam computed tomography （CBCT） using the flat-panel detector, imperfect or defect detector elements cause ring artifacts due to the none-uniformity of their X-ray response. They often distur...In high-resolution cone-beam computed tomography （CBCT） using the flat-panel detector, imperfect or defect detector elements cause ring artifacts due to the none-uniformity of their X-ray response. They often disturb the image quality. A dedicated fitting correction method for high-resolution micro-CT is presented. The method converts each elementary X-ray response curve to an average one, and eliminates response inconsistency among pixels. Other factors of the method are discussed, such as the correction factor variability by different sampling frames and nonlinear factors over the whole spectrum. Results show that the noise and artifacts are both reduced in reconstructed images展开更多
To improve spectral X-ray CT reconstructed image quality, the energy-weighted reconstructed image xbins^W and the separable paraboloidal surrogates（SPS） algorithm are proposed for the prior image constrained compres...To improve spectral X-ray CT reconstructed image quality, the energy-weighted reconstructed image xbins^W and the separable paraboloidal surrogates（SPS） algorithm are proposed for the prior image constrained compressed sensing（PICCS）-based spectral X-ray CT image reconstruction. The PICCS-based image reconstruction takes advantage of the compressed sensing theory, a prior image and an optimization algorithm to improve the image quality of CT reconstructions.To evaluate the performance of the proposed method, three optimization algorithms and three prior images are employed and compared in terms of reconstruction accuracy and noise characteristics of the reconstructed images in each energy bin.The experimental simulation results show that the image xbins^W is the best as the prior image in general with respect to the three optimization algorithms; and the SPS algorithm offers the best performance for the simulated phantom with respect to the three prior images. Compared with filtered back-projection（FBP）, the PICCS via the SPS algorithm and xbins^W as the prior image can offer the noise reduction in the reconstructed images up to 80. 46%, 82. 51%, 88. 08% in each energy bin,respectively. M eanwhile, the root-mean-squared error in each energy bin is decreased by 15. 02%, 18. 15%, 34. 11% and the correlation coefficient is increased by 9. 98%, 11. 38%,15. 94%, respectively.展开更多
Muon tomography is a novel method for the non-destructive imaging of materials based on muon rays,which are highly penetrating in natural background radiation.Currently,the most commonly used imaging methods include m...Muon tomography is a novel method for the non-destructive imaging of materials based on muon rays,which are highly penetrating in natural background radiation.Currently,the most commonly used imaging methods include muon radiography and muon tomography.A previously studied method known as coinciding muon trajectory density tomography,which utilizes muonic secondary particles,is proposed to image low and medium atomic number(Z)materials.However,scattering tomography is mostly used to image high-Z materials,and coinciding muon trajectory density tomography exhibits a hollow phenomenon in the imaging results owing to the self-absorption effect.To address the shortcomings of the individual imaging methods,hybrid model tomography combining scattering tomography and coinciding muon trajectory density tomography is proposed and verified.In addition,the peak signal-to-noise ratio was introduced to quantitatively analyze the image quality.Different imaging models were simulated using the Geant4 toolkit to confirm the advantages of this innovative method.The simulation results showed that hybrid model tomography can image centimeter-scale materials with low,medium,and high Z simultaneously.For high-Z materials with similar atomic numbers,this method can clearly distinguish those with apparent differences in density.According to the peak signal-to-noise ratio of the analysis,the reconstructed image quality of the new method was significantly higher than that of the individual imaging methods.This study provides a reliable approach to the compatibility of scattering tomography and coinciding muon trajectory density tomography.展开更多
Linear scan computed tomography （LCT） is of great benefit to online industrial scanning and security inspection due to its characteristics of straight-line source trajectory and high scanning speed. However, in prac...Linear scan computed tomography （LCT） is of great benefit to online industrial scanning and security inspection due to its characteristics of straight-line source trajectory and high scanning speed. However, in practical applications of LCT, there are challenges to image reconstruction due to limited-angle and insufficient data. In this paper, a new reconstruction algorithm based on total-variation （TV） minimization is developed to reconstruct images from limited-angle and insufficient data in LCT. The main idea of our approach is to reformulate a TV problem as a linear equality constrained problem where the objective function is separable, and then minimize its augmented Lagrangian function by using alternating direction method （ADM） to solve subproblems. The proposed method is robust and efficient in the task of reconstruction by showing the convergence of ADM. The numerical simulations and real data reconstructions show that the proposed reconstruction method brings reasonable performance and outperforms some previous ones when applied to an LCT imaging problem.展开更多
With the development of the compressive sensing theory, the image reconstruction from the projections viewed in limited angles is one of the hot problems in the research of computed tomography technology. This paper d...With the development of the compressive sensing theory, the image reconstruction from the projections viewed in limited angles is one of the hot problems in the research of computed tomography technology. This paper develops an iterative algorithm for image reconstruction, which can fit the most cases. This method gives an image reconstruction flow with the difference image vector, which is based on the concept that the difference image vector between the reconstructed and the reference image is sparse enough. Then the l1-norm minimization method is used to reconstruct the difference vector to recover the image for flat subjects in limited angles. The algorithm has been tested with a thin planar phantom and a real object in limited-view projection data. Moreover, all the studies showed the satisfactory results in accuracy at a rather high reconstruction speed.展开更多
Electrical capacitance tomography(ECT)has been applied to two-phase flow measurement in recent years.Image reconstruction algorithms play an important role in the successful applications of ECT.To solve the ill-posed ...Electrical capacitance tomography(ECT)has been applied to two-phase flow measurement in recent years.Image reconstruction algorithms play an important role in the successful applications of ECT.To solve the ill-posed and nonlinear inverse problem of ECT image reconstruction,a new ECT image reconstruction method based on fast linearized alternating direction method of multipliers(FLADMM)is proposed in this paper.On the basis of theoretical analysis of compressed sensing(CS),the data acquisition of ECT is regarded as a linear measurement process of permittivity distribution signal of pipe section.A new measurement matrix is designed and L1 regularization method is used to convert ECT inverse problem to a convex relaxation problem which contains prior knowledge.A new fast alternating direction method of multipliers which contained linearized idea is employed to minimize the objective function.Simulation data and experimental results indicate that compared with other methods,the quality and speed of reconstructed images are markedly improved.Also,the dynamic experimental results indicate that the proposed algorithm can ful fill the real-time requirement of ECT systems in the application.展开更多
As-built building information model （BIM） is an urgent need of the architecture, engineering, construction and facilities management （AEC/FM） community. However, its creation procedure is still labor-intensive and...As-built building information model （BIM） is an urgent need of the architecture, engineering, construction and facilities management （AEC/FM） community. However, its creation procedure is still labor-intensive and far from maturity. Taking advantage of prevalence of digital cameras and the development of advanced computer vision technology, the paper proposes to reconstruct a building facade and recognize its surface materials from images taken from various points of view. These can serve as initial steps towards automatic generation of as-built BIM. Specifically, 3D point clouds are generated from multiple images using structure from motion method and then segmented into planar components, which are further recognized as different structural components through knowledge based reasoning. Windows are detected through a multilayered complementary strategy by combining detection results from every semantic layer. A novel machine learning based 3D material recognition strategy is presented. Binary classifiers are trained through support vector machines. Material type at a given 3D location is predicted by all its corresponding 2D feature points. Experimental results from three existing buildings validate the proposed system.展开更多
A super-resolution reconstruction approach of （SVD） technique was presented, and its performance was radar image using an adaptive-threshold singular value decomposition analyzed, compared and assessed detailedly. F...A super-resolution reconstruction approach of （SVD） technique was presented, and its performance was radar image using an adaptive-threshold singular value decomposition analyzed, compared and assessed detailedly. First, radar imaging model and super-resolution reconstruction mechanism were outlined. Then, the adaptive-threshold SVD super-resolution algorithm, and its two key aspects, namely the determination method of point spread function （PSF） matrix T and the selection scheme of singular value threshold, were presented. Finally, the super-resolution algorithm was demonstrated successfully using the measured synthetic-aperture radar （SAR） images, and a Monte Carlo assessment was carried out to evaluate the performance of the algorithm by using the input/output signal-to-noise ratio （SNR）. Five versions of SVD algorithms, namely 1 ） using all singular values, 2） using the top 80% singular values, 3） using the top 50% singular values, 4） using the top 20% singular values and 5） using singular values s such that S2≥/max（s2）/rinsNR were tested. The experimental results indicate that when the singular value threshold is set as Smax/（rinSNR）1/2, the super-resolution algorithm provides a good compromise between too much noise and too much bias and has good reconstruction results.展开更多
Electrical capacitance tomography(ECT) is a non-invasive imaging technique that aims at visualizing the cross-sectional permittivity distribution and phase distribution of solid/gas two-phase flow based on the measure...Electrical capacitance tomography(ECT) is a non-invasive imaging technique that aims at visualizing the cross-sectional permittivity distribution and phase distribution of solid/gas two-phase flow based on the measured capacitance.To solve the nonlinear and ill-posed inverse problem:image reconstruction of ECT system,this paper proposed a new image reconstruction method based on improved radial basis function(RBF) neural network combined with adaptive wavelet image enhancement.Firstly,an improved RBF network was applied to establish the mapping model between the reconstruction image pixels and the capacitance values measured.Then,for better image quality,adaptive wavelet image enhancement technique was emphatically analyzed and studied,which belongs to a space-frequency analysis method and is suitable for image feature-enhanced.Through multi-level wavelet decomposition,edge points of the image produced from RBF network can be determined based on the neighborhood property of each sub-band;noise distribution in the space-frequency domain can be estimated based on statistical characteristics;after that a self-adaptive edge enhancement gain can be constructed.Finally,the image is reconstructed with adjusting wavelet coefficients.In this paper,a 12-electrode ECT system and a pneumatic conveying platform were built up to verify this image reconstruction algorithm.Experimental results demonstrated that adaptive wavelet image enhancement technique effectively implemented edge detection and image enhancement,and the improved RBF network and adaptive wavelet image enhancement hybrid algorithm greatly improved the quality of reconstructed image of solid/gas two-phase flow [pulverized coal(PC)/air].展开更多
The evaluation approach to the accuracy of the image feature descriptors plays an important role in image feature extraction. We point out that the image shape feature can be described by the Zernike moments set while...The evaluation approach to the accuracy of the image feature descriptors plays an important role in image feature extraction. We point out that the image shape feature can be described by the Zernike moments set while briefly introducing the basic concept of the Zernike moment. After talking about the image reconstruction technique based on the inverse transformation of Zernike moment, the evaluation approach to the accuracy of the Zernike moments shape feature via the dissimilarity degree and the reconstruction ratio between the original image and the reconstructed image is proposed. The experiment results demonstrate the feasibility of this evaluation approach to image Zernike moments shape feature.展开更多
The Computer Tomography(CT)method is used for remote sensing the Earth’s plasmasphere.One challenge for image reconstruction is insufficient projection data,mainly caused by limited projection angles.In this study,we...The Computer Tomography(CT)method is used for remote sensing the Earth’s plasmasphere.One challenge for image reconstruction is insufficient projection data,mainly caused by limited projection angles.In this study,we apply the Algebraic Reconstruction Technique(ART)and the minimization of the image Total Variation(TV)method,with a combination of priori knowledge of north–south symmetry,to reconstruct plasmaspheric He+density from simulated EUV images.The results demonstrate that incorporating priori assumption can be particularly useful when the projection data is insufficient.This method has good performance even with a projection angle of less than 150 degrees.The method of our study is expected to have applications in the Soft X-ray Imager(SXI)reconstruction for the Solar wind–Magnetosphere–Ionosphere Link Explorer(SMILE)mission.展开更多
Proton computed tomography(CT)has a distinct practical significance in clinical applications.It eliminates 3–5%errors caused by the transformation of Hounsfield unit(HU)to relative stopping power(RSP)values when usin...Proton computed tomography(CT)has a distinct practical significance in clinical applications.It eliminates 3–5%errors caused by the transformation of Hounsfield unit(HU)to relative stopping power(RSP)values when using X-ray CT for positioning and treatment planning systems(TPSs).Following the development of FLASH proton therapy,there are increased requirements for accurate and rapid positioning in TPSs.Thus,a new rapid proton CT imaging mode is proposed based on sparsely sampled projections.The proton beam was boosted to 350 MeV by a compact proton linear accelerator(LINAC).In this study,the comparisons of the proton scattering with the energy of 350 MeV and 230 MeV are conducted based on GEANT4 simulations.As the sparsely sampled information associated with beam acquisitions at 12 angles is not enough for reconstruction,X-ray CT is used as a prior image.The RSP map generated by converting the X-ray CT was constructed based on Monte Carlo simulations.Considering the estimation of the most likely path(MLP),the prior image-constrained compressed sensing(PICCS)algorithm is used to reconstruct images from two different phantoms using sparse proton projections of 350 MeV parallel proton beam.The results show that it is feasible to realize the proton image reconstruction with the rapid proton CT imaging proposed in this paper.It can produce RSP maps with much higher accuracy for TPSs and fast positioning to achieve ultra-fast imaging for real-time image-guided radiotherapy(IGRT)in clinical proton therapy applications.展开更多
Structured illumination microscopy(SIM)achieves super-resolution(SR)by modulating the high-frequency information of the sample into the passband of the optical system and subsequent image reconstruction.The traditiona...Structured illumination microscopy(SIM)achieves super-resolution(SR)by modulating the high-frequency information of the sample into the passband of the optical system and subsequent image reconstruction.The traditional Wiener-filtering-based reconstruction algorithm operates in the Fourier domain,it requires prior knowledge of the sinusoidal illumination patterns which makes the time-consuming procedure of parameter estimation to raw datasets necessary,besides,the parameter estimation is sensitive to noise or aberration-induced pattern distortion which leads to reconstruction artifacts.Here,we propose a spatial-domain image reconstruction method that does not require parameter estimation but calculates patterns from raw datasets,and a reconstructed image can be obtained just by calculating the spatial covariance of differential calculated patterns and differential filtered datasets(the notch filtering operation is performed to the raw datasets for attenuating and compensating the optical transfer function(OTF)).Experiments on reconstructing raw datasets including nonbiological,biological,and simulated samples demonstrate that our method has SR capability,high reconstruction speed,and high robustness to aberration and noise.展开更多
文摘Image super-resolution reconstruction technology is currently widely used in medical imaging,video surveillance,and industrial quality inspection.It not only enhances image quality but also improves details and visual perception,significantly increasing the utility of low-resolution images.In this study,an improved image superresolution reconstruction model based on Generative Adversarial Networks(SRGAN)was proposed.This model introduced a channel and spatial attention mechanism(CSAB)in the generator,allowing it to effectively leverage the information from the input image to enhance feature representations and capture important details.The discriminator was designed with an improved PatchGAN architecture,which more accurately captured local details and texture information of the image.With these enhanced generator and discriminator architectures and an optimized loss function design,this method demonstrated superior performance in image quality assessment metrics.Experimental results showed that this model outperforms traditional methods,presenting more detailed and realistic image details in the visual effects.
文摘In agricultural production,crop images are commonly used for the classification and identification of various crops.However,several challenges arise,including low image clarity,elevated noise levels,low accuracy,and poor robustness of existing classification models.To address these issues,this research proposes an innovative crop image classification model named Lap-FEHRNet,which integrates a Laplacian Pyramid Super Resolution Network(LapSRN)with a feature enhancement high-resolution network based on attention mechanisms(FEHRNet).To mitigate noise interference,this research incorporates the LapSRN network,which utilizes a Laplacian pyramid structure to extract multi-level feature details from low-resolution images through a systematic layer-by-layer amplification and pixel detail superposition process.This gradual reconstruction enhances the high-frequency information of the image,enabling super-resolution reconstruction of low-quality images.To obtain a broader range of comprehensive and diverse features,this research employs the FEHRNetmodel for both deep and shallow feature extraction.This approach results in features that encapsulate multi-scale information and integrate both deep and shallow insights.To effectively fuse these complementary features,this research introduces an attention mechanism during the feature enhancement stage.This mechanism highlights important regions within the image,assigning greater weights to salient features and resulting in a more comprehensive and effective image feature representation.Consequently,the accuracy of image classification is significantly improved.Experimental results demonstrate that the Lap-FEHRNetmodel achieves impressive classification accuracies of 98.8%on the crop classification dataset and 98.57%on the rice leaf disease dataset,underscoring the model’s outstanding accuracy,robustness,and generalization capability.
基金supported by the Interdisciplinary project of Dalian University DLUXK-2023-ZD-001.
文摘As a form of discrete representation learning,Vector Quantized Variational Autoencoders(VQ-VAE)have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capacity.However,existing VQ-VAEs often perform quantization in the spatial domain,ignoring global structural information and potentially suffering from codebook collapse and information coupling issues.This paper proposes a frequency quantized variational autoencoder(FQ-VAE)to address these issues.The proposed method transforms image features into linear combinations in the frequency domain using a 2D fast Fourier transform(2D-FFT)and performs adaptive quantization on these frequency components to preserve image’s global relationships.The codebook is dynamically optimized to avoid collapse and information coupling issue by considering the usage frequency and dependency of code vectors.Furthermore,we introduce a post-processing module based on graph convolutional networks to further improve reconstruction quality.Experimental results on four public datasets demonstrate that the proposed method outperforms state-of-the-art approaches in terms of Structural Similarity Index(SSIM),Learned Perceptual Image Patch Similarity(LPIPS),and Reconstruction Fréchet Inception Distance(rFID).In the experiments on the CIFAR-10 dataset,compared to the baselinemethod VQ-VAE,the proposedmethod improves the abovemetrics by 4.9%,36.4%,and 52.8%,respectively.
基金supported in part by the National Natural Science Foundation of China(62101278,62001379,62271023)Beijing Natural Science Foundation(7242269).
文摘Deep learning(DL)-based image reconstruction methods have garnered increasing interest in the last few years.Numerous studies demonstrate that DL-based reconstruction methods function admirably in optical tomographic imaging techniques,such as bioluminescence tomography(BLT).Nevertheless,nearly every existing DL-based method utilizes an explicit neural representation for the reconstruction problem,which either consumes much memory space or requires various complicated computations.In this paper,we present a neural field(NF)-based image reconstruction scheme for BLT that uses an implicit neural representation.The proposed NFbased method establishes a transformation between the coordinate of an arbitrary spatial point and the source value of the point with a relatively light-weight multilayer perceptron,which has remarkable computational efficiency.Another simple neural network composed of two fully connected layers and a 1D convolutional layer is used to generate the neural features.Results of simulations and experiments show that the proposed NF-based method has similar performance to the photon density complement network and the two-stage network,while consuming fewer floating point operations with fewer model parameters.
基金National Natural Science Foundation of China,Grant/Award Number:62176124。
文摘Person Image Synthesis has been widely used in fashion with extensive application scenarios.The point of this task is how to synthesise person image from a single source image under arbitrary poses.Prior methods generate the person image with target pose well;however,they fail to preserve the fine style details of the source image.To address this problem,a robust style injection(RSI)model is proposed,which is a coarse-to-fine framework to synthesise target the person image.RSI develops a simple and efficient cross-attention based module to fuse the features of both source semantic styles and target pose for achieving the coarse aligned features.The adaptive instance normalisation is employed to enhance the aligned features in conjunction with source semantic styles.Subsequently,source semantic styles are further injected into the positional normalisation scheme to avoid the fine style details erosion caused by massive convolution.In training losses,optimal transport theory in the form of energy distance is introduced to constrain data distribution to refine the texture style details.Additionally,the authors’model is capable of editing the shape and texture of garments to the target style separately.The experiments demonstrate that the authors’RSI achieves better performance over the state-of-art methods.
基金National Natural Science Foundation of China(Grant Nos.62005049 and 62072110)Natural Science Foundation of Fujian Province(Grant No.2020J01451).
文摘Accurate segmentation of camouflage objects in aerial imagery is vital for improving the efficiency of UAV-based reconnaissance and rescue missions.However,camouflage object segmentation is increasingly challenging due to advances in both camouflage materials and biological mimicry.Although multispectral-RGB based technology shows promise,conventional dual-aperture multispectral-RGB imaging systems are constrained by imprecise and time-consuming registration and fusion across different modalities,limiting their performance.Here,we propose the Reconstructed Multispectral-RGB Fusion Network(RMRF-Net),which reconstructs RGB images into multispectral ones,enabling efficient multimodal segmentation using only an RGB camera.Specifically,RMRF-Net employs a divergentsimilarity feature correction strategy to minimize reconstruction errors and includes an efficient boundary-aware decoder to enhance object contours.Notably,we establish the first real-world aerial multispectral-RGB semantic segmentation of camouflage objects dataset,including 11 object categories.Experimental results demonstrate that RMRF-Net outperforms existing methods,achieving 17.38 FPS on the NVIDIA Jetson AGX Orin,with only a 0.96%drop in mIoU compared to the RTX 3090,showing its practical applicability in multimodal remote sensing.
文摘A new method to accelerate the convergent rate of the space-alternatinggeneralized expectation-maximization (SAGE) algorithm is proposed. The new rescaled block-iterativeSAGE (RBI-SAGE) algorithm combines the RBI algorithm with the SAGE algorithm for PET imagereconstruction. In the new approach, the projection data is partitioned into disjoint blocks; eachiteration step involves only one of these blocks. SAGE updates the parameters sequentially in eachblock. In experiments, the RBI-SAGE algorithm and classical SAGE algorithm are compared in theapplication on positron emission tomography (PET) image reconstruction. Simulation results show thatRBI-SAGE has better performance than SAGE in both convergence and image quality.
基金Supported by the National Basic Research Program of China ("973"Program)(2006CB601201)~~
文摘In high-resolution cone-beam computed tomography （CBCT） using the flat-panel detector, imperfect or defect detector elements cause ring artifacts due to the none-uniformity of their X-ray response. They often disturb the image quality. A dedicated fitting correction method for high-resolution micro-CT is presented. The method converts each elementary X-ray response curve to an average one, and eliminates response inconsistency among pixels. Other factors of the method are discussed, such as the correction factor variability by different sampling frames and nonlinear factors over the whole spectrum. Results show that the noise and artifacts are both reduced in reconstructed images
基金The National Natural Science Foundation of China(No.51575256)the Fundamental Research Funds for the Central Universities(No.NP2015101,XZA16003)the Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)
文摘To improve spectral X-ray CT reconstructed image quality, the energy-weighted reconstructed image xbins^W and the separable paraboloidal surrogates（SPS） algorithm are proposed for the prior image constrained compressed sensing（PICCS）-based spectral X-ray CT image reconstruction. The PICCS-based image reconstruction takes advantage of the compressed sensing theory, a prior image and an optimization algorithm to improve the image quality of CT reconstructions.To evaluate the performance of the proposed method, three optimization algorithms and three prior images are employed and compared in terms of reconstruction accuracy and noise characteristics of the reconstructed images in each energy bin.The experimental simulation results show that the image xbins^W is the best as the prior image in general with respect to the three optimization algorithms; and the SPS algorithm offers the best performance for the simulated phantom with respect to the three prior images. Compared with filtered back-projection（FBP）, the PICCS via the SPS algorithm and xbins^W as the prior image can offer the noise reduction in the reconstructed images up to 80. 46%, 82. 51%, 88. 08% in each energy bin,respectively. M eanwhile, the root-mean-squared error in each energy bin is decreased by 15. 02%, 18. 15%, 34. 11% and the correlation coefficient is increased by 9. 98%, 11. 38%,15. 94%, respectively.
基金supported by the National Natural Science Foundation of China(No.11875163)Natural Science Foundation of Hunan Province(Nos.2021JJ20006 and 2021JJ40444)+1 种基金Ministry of Science and Technology of China(No.2020YFE0202001)Department of Education of Hunan Province(Nos.19B488 and 21A0281)。
文摘Muon tomography is a novel method for the non-destructive imaging of materials based on muon rays,which are highly penetrating in natural background radiation.Currently,the most commonly used imaging methods include muon radiography and muon tomography.A previously studied method known as coinciding muon trajectory density tomography,which utilizes muonic secondary particles,is proposed to image low and medium atomic number(Z)materials.However,scattering tomography is mostly used to image high-Z materials,and coinciding muon trajectory density tomography exhibits a hollow phenomenon in the imaging results owing to the self-absorption effect.To address the shortcomings of the individual imaging methods,hybrid model tomography combining scattering tomography and coinciding muon trajectory density tomography is proposed and verified.In addition,the peak signal-to-noise ratio was introduced to quantitatively analyze the image quality.Different imaging models were simulated using the Geant4 toolkit to confirm the advantages of this innovative method.The simulation results showed that hybrid model tomography can image centimeter-scale materials with low,medium,and high Z simultaneously.For high-Z materials with similar atomic numbers,this method can clearly distinguish those with apparent differences in density.According to the peak signal-to-noise ratio of the analysis,the reconstructed image quality of the new method was significantly higher than that of the individual imaging methods.This study provides a reliable approach to the compatibility of scattering tomography and coinciding muon trajectory density tomography.
基金the National High Technology Research and Development Program of China(Grant No.2012AA011603)
文摘Linear scan computed tomography （LCT） is of great benefit to online industrial scanning and security inspection due to its characteristics of straight-line source trajectory and high scanning speed. However, in practical applications of LCT, there are challenges to image reconstruction due to limited-angle and insufficient data. In this paper, a new reconstruction algorithm based on total-variation （TV） minimization is developed to reconstruct images from limited-angle and insufficient data in LCT. The main idea of our approach is to reformulate a TV problem as a linear equality constrained problem where the objective function is separable, and then minimize its augmented Lagrangian function by using alternating direction method （ADM） to solve subproblems. The proposed method is robust and efficient in the task of reconstruction by showing the convergence of ADM. The numerical simulations and real data reconstructions show that the proposed reconstruction method brings reasonable performance and outperforms some previous ones when applied to an LCT imaging problem.
基金Project supported by the National Basic Research Program of China(Grant No.2006CB7057005)the National High Technology Research and Development Program of China(Grant No.2009AA012200)the National Natural Science Foundation of China (Grant No.60672104)
文摘With the development of the compressive sensing theory, the image reconstruction from the projections viewed in limited angles is one of the hot problems in the research of computed tomography technology. This paper develops an iterative algorithm for image reconstruction, which can fit the most cases. This method gives an image reconstruction flow with the difference image vector, which is based on the concept that the difference image vector between the reconstructed and the reference image is sparse enough. Then the l1-norm minimization method is used to reconstruct the difference vector to recover the image for flat subjects in limited angles. The algorithm has been tested with a thin planar phantom and a real object in limited-view projection data. Moreover, all the studies showed the satisfactory results in accuracy at a rather high reconstruction speed.
基金Supported by the National Natural Science Foundation of China(61203021)the Key Science and Technology Program of Liaoning Province(2011216011)+1 种基金the Natural Science Foundation of Liaoning Province(2013020024)the Program for Liaoning Excellent Talents in Universities(LJQ2015061)
文摘Electrical capacitance tomography(ECT)has been applied to two-phase flow measurement in recent years.Image reconstruction algorithms play an important role in the successful applications of ECT.To solve the ill-posed and nonlinear inverse problem of ECT image reconstruction,a new ECT image reconstruction method based on fast linearized alternating direction method of multipliers(FLADMM)is proposed in this paper.On the basis of theoretical analysis of compressed sensing(CS),the data acquisition of ECT is regarded as a linear measurement process of permittivity distribution signal of pipe section.A new measurement matrix is designed and L1 regularization method is used to convert ECT inverse problem to a convex relaxation problem which contains prior knowledge.A new fast alternating direction method of multipliers which contained linearized idea is employed to minimize the objective function.Simulation data and experimental results indicate that compared with other methods,the quality and speed of reconstructed images are markedly improved.Also,the dynamic experimental results indicate that the proposed algorithm can ful fill the real-time requirement of ECT systems in the application.
基金supported by National Natural Science Foundation of China(No.51208425)Research Foundation of Northwestern Polytechnical University(No.JCY20130127)
文摘As-built building information model （BIM） is an urgent need of the architecture, engineering, construction and facilities management （AEC/FM） community. However, its creation procedure is still labor-intensive and far from maturity. Taking advantage of prevalence of digital cameras and the development of advanced computer vision technology, the paper proposes to reconstruct a building facade and recognize its surface materials from images taken from various points of view. These can serve as initial steps towards automatic generation of as-built BIM. Specifically, 3D point clouds are generated from multiple images using structure from motion method and then segmented into planar components, which are further recognized as different structural components through knowledge based reasoning. Windows are detected through a multilayered complementary strategy by combining detection results from every semantic layer. A novel machine learning based 3D material recognition strategy is presented. Binary classifiers are trained through support vector machines. Material type at a given 3D location is predicted by all its corresponding 2D feature points. Experimental results from three existing buildings validate the proposed system.
基金Project(2008041001) supported by the Academician Foundation of China Project(N0601-041) supported by the General Armament Department Science Foundation of China
文摘A super-resolution reconstruction approach of （SVD） technique was presented, and its performance was radar image using an adaptive-threshold singular value decomposition analyzed, compared and assessed detailedly. First, radar imaging model and super-resolution reconstruction mechanism were outlined. Then, the adaptive-threshold SVD super-resolution algorithm, and its two key aspects, namely the determination method of point spread function （PSF） matrix T and the selection scheme of singular value threshold, were presented. Finally, the super-resolution algorithm was demonstrated successfully using the measured synthetic-aperture radar （SAR） images, and a Monte Carlo assessment was carried out to evaluate the performance of the algorithm by using the input/output signal-to-noise ratio （SNR）. Five versions of SVD algorithms, namely 1 ） using all singular values, 2） using the top 80% singular values, 3） using the top 50% singular values, 4） using the top 20% singular values and 5） using singular values s such that S2≥/max（s2）/rinsNR were tested. The experimental results indicate that when the singular value threshold is set as Smax/（rinSNR）1/2, the super-resolution algorithm provides a good compromise between too much noise and too much bias and has good reconstruction results.
基金Supported by the National Natural Science Foundation of China (50777049,51177120)the National High Technology Research and Development Program of China (2009AA04Z130)the RCUK’s Energy Programme (EP/F061307/1)
文摘Electrical capacitance tomography(ECT) is a non-invasive imaging technique that aims at visualizing the cross-sectional permittivity distribution and phase distribution of solid/gas two-phase flow based on the measured capacitance.To solve the nonlinear and ill-posed inverse problem:image reconstruction of ECT system,this paper proposed a new image reconstruction method based on improved radial basis function(RBF) neural network combined with adaptive wavelet image enhancement.Firstly,an improved RBF network was applied to establish the mapping model between the reconstruction image pixels and the capacitance values measured.Then,for better image quality,adaptive wavelet image enhancement technique was emphatically analyzed and studied,which belongs to a space-frequency analysis method and is suitable for image feature-enhanced.Through multi-level wavelet decomposition,edge points of the image produced from RBF network can be determined based on the neighborhood property of each sub-band;noise distribution in the space-frequency domain can be estimated based on statistical characteristics;after that a self-adaptive edge enhancement gain can be constructed.Finally,the image is reconstructed with adjusting wavelet coefficients.In this paper,a 12-electrode ECT system and a pneumatic conveying platform were built up to verify this image reconstruction algorithm.Experimental results demonstrated that adaptive wavelet image enhancement technique effectively implemented edge detection and image enhancement,and the improved RBF network and adaptive wavelet image enhancement hybrid algorithm greatly improved the quality of reconstructed image of solid/gas two-phase flow [pulverized coal(PC)/air].
文摘The evaluation approach to the accuracy of the image feature descriptors plays an important role in image feature extraction. We point out that the image shape feature can be described by the Zernike moments set while briefly introducing the basic concept of the Zernike moment. After talking about the image reconstruction technique based on the inverse transformation of Zernike moment, the evaluation approach to the accuracy of the Zernike moments shape feature via the dissimilarity degree and the reconstruction ratio between the original image and the reconstructed image is proposed. The experiment results demonstrate the feasibility of this evaluation approach to image Zernike moments shape feature.
基金supported by the National Natural Science Foundation of China(Grant Nos.41904148,41731070,41874175)in part by the Strategic Priority Program on Space Science,Chinese Academy of Sciences(Grant Nos.XDA15017000,XDA15350201,XDA15052500).
文摘The Computer Tomography(CT)method is used for remote sensing the Earth’s plasmasphere.One challenge for image reconstruction is insufficient projection data,mainly caused by limited projection angles.In this study,we apply the Algebraic Reconstruction Technique(ART)and the minimization of the image Total Variation(TV)method,with a combination of priori knowledge of north–south symmetry,to reconstruct plasmaspheric He+density from simulated EUV images.The results demonstrate that incorporating priori assumption can be particularly useful when the projection data is insufficient.This method has good performance even with a projection angle of less than 150 degrees.The method of our study is expected to have applications in the Soft X-ray Imager(SXI)reconstruction for the Solar wind–Magnetosphere–Ionosphere Link Explorer(SMILE)mission.
基金supported by the Research collaboration on Thailand’s new synchrotron light source facility(SPS-II)(No.ANSO-CR-KP-2020-16).
文摘Proton computed tomography(CT)has a distinct practical significance in clinical applications.It eliminates 3–5%errors caused by the transformation of Hounsfield unit(HU)to relative stopping power(RSP)values when using X-ray CT for positioning and treatment planning systems(TPSs).Following the development of FLASH proton therapy,there are increased requirements for accurate and rapid positioning in TPSs.Thus,a new rapid proton CT imaging mode is proposed based on sparsely sampled projections.The proton beam was boosted to 350 MeV by a compact proton linear accelerator(LINAC).In this study,the comparisons of the proton scattering with the energy of 350 MeV and 230 MeV are conducted based on GEANT4 simulations.As the sparsely sampled information associated with beam acquisitions at 12 angles is not enough for reconstruction,X-ray CT is used as a prior image.The RSP map generated by converting the X-ray CT was constructed based on Monte Carlo simulations.Considering the estimation of the most likely path(MLP),the prior image-constrained compressed sensing(PICCS)algorithm is used to reconstruct images from two different phantoms using sparse proton projections of 350 MeV parallel proton beam.The results show that it is feasible to realize the proton image reconstruction with the rapid proton CT imaging proposed in this paper.It can produce RSP maps with much higher accuracy for TPSs and fast positioning to achieve ultra-fast imaging for real-time image-guided radiotherapy(IGRT)in clinical proton therapy applications.
基金funded by the National Natural Science Foundation of China(62125504,61827825,and 31901059)Zhejiang Provincial Ten Thousand Plan for Young Top Talents(2020R52001)Open Project Program of Wuhan National Laboratory for Optoelectronics(2021WNLOKF007).
文摘Structured illumination microscopy(SIM)achieves super-resolution(SR)by modulating the high-frequency information of the sample into the passband of the optical system and subsequent image reconstruction.The traditional Wiener-filtering-based reconstruction algorithm operates in the Fourier domain,it requires prior knowledge of the sinusoidal illumination patterns which makes the time-consuming procedure of parameter estimation to raw datasets necessary,besides,the parameter estimation is sensitive to noise or aberration-induced pattern distortion which leads to reconstruction artifacts.Here,we propose a spatial-domain image reconstruction method that does not require parameter estimation but calculates patterns from raw datasets,and a reconstructed image can be obtained just by calculating the spatial covariance of differential calculated patterns and differential filtered datasets(the notch filtering operation is performed to the raw datasets for attenuating and compensating the optical transfer function(OTF)).Experiments on reconstructing raw datasets including nonbiological,biological,and simulated samples demonstrate that our method has SR capability,high reconstruction speed,and high robustness to aberration and noise.
