This paper proposes a subpixel transformation method to correct Keystone and Smile distortions in fiber spectral images from the Fiber Arrayed Solar Optical Telescope.These distortions affect the spatial and spectral ...This paper proposes a subpixel transformation method to correct Keystone and Smile distortions in fiber spectral images from the Fiber Arrayed Solar Optical Telescope.These distortions affect the spatial and spectral positions,degrading resolution and accuracy.To correct Keystone distortion,we use a local summation and peak-finding method to locate central horizontal coordinates,calculate shifting values,and straighten the curves.For Smile distortion,we use quartic polynomial fitting based on absorption lines at different wavelengths.This technique preserves subpixel components,redistributes pixel values,and interpolates non-fiber portions,rectifying the spectra for accurate analysis.The method can also be applied to other astronomical projects like Large Sky Area Multi-Object Fiber Spectroscopic Telescope,enhancing the accuracy and reliability of spectral data in various astronomical studies.展开更多
Deconvolution in radio interferometry faces challenges due to incomplete sampling of the visibilities in the spatial frequency domain caused by a limited number of antenna baselines,resulting in an ill-posed inverse p...Deconvolution in radio interferometry faces challenges due to incomplete sampling of the visibilities in the spatial frequency domain caused by a limited number of antenna baselines,resulting in an ill-posed inverse problem.Reconstructing dirty images into clean ones is crucial for subsequent scientific analysis.To address these challenges,we propose a U-Net based method that extracts high-level information from the dirty image and reconstructs a clean image by effectively reducing artifacts and sidelobes.The U-Net architecture,consisting of an encoder-decoder structure and skip connections,facilitates the flow of information and preserves spatial details.Using simulated data of radio galaxies,we train our model and evaluate its performance on the testing set.Compared with the CLEAN method and the visibility and image conditioned denoising diffusion probabilistic model,our proposed model can effectively reconstruct both extended sources and faint point sources with higher values in the structural similarity index measure and the peak signal-to-noise ratio.Furthermore,we investigate the impact of noise on the model performance,demonstrating its robustness under varying noise levels.展开更多
Imaging is an important method for astronomy research.In practice,original images acquired by a telescope are often convolved and blurred by the point-spread function(PSF),which is a very unfavorable situation for man...Imaging is an important method for astronomy research.In practice,original images acquired by a telescope are often convolved and blurred by the point-spread function(PSF),which is a very unfavorable situation for many scientific studies including astronomy.This paper introduced a single equation iterative method for solving complex linear equations,and this method can deconvolute dirty images,eliminate the effects of the PSF well.With different PSFs,this algorithm shows very good results in deconvolution.Also,with a giant PSF of aperture synthesis imaging,this algorithm improves the peak signal-to-noise ratio and structural similarity of the dirty images by 41.0%and 33.9%on average.In addition,this paper proves that the algorithm can deconvolute the dirty image by making full use of the information of each pixel in the image,even if the dirty image has salt and pepper noise or even lost areas;by its excellent properties of flexible operation to a single pixel,all these bad situations can be dealt with and the image can be restored.展开更多
Aperture photometry is a fundamental technique widely used to obtain high-precision light curves in optical survey projects like Tianyu.However,its effectiveness is limited in crowded fields,and the choice of aperture...Aperture photometry is a fundamental technique widely used to obtain high-precision light curves in optical survey projects like Tianyu.However,its effectiveness is limited in crowded fields,and the choice of aperture size critically impacts photometric precision.To address these challenges,we propose DeepAP,an efficient and accurate two-stage deep learning framework for aperture photometry.Specifically,for a given source,we first train a Vision Transformer(ViT)model to assess its feasibility of aperture photometry.We then train the Residual Neural Network(ResNet)to predict its optimal aperture size.For aperture photometry feasibility assessment,the ViT model yields an ROC AUC value of 0.96,and achieves a precision of 0.974,a recall of 0.930,and an F1 score of 0.952 on the test set.For aperture size prediction,the ResNet model effectively mitigates biases inherent in classical growth curve methods by adaptively selecting apertures appropriate for sources of varying brightness,thereby enhancing the signal-to-noise ratio(SNR)across a wide range of targets.Meanwhile,some samples in the test set have a higher SNR than those obtained by exhaustive aperture size enumeration because of the finer granularity of aperture size estimation.By integrating ResNet with the ViT network,the DeepAP framework achieves a median total processing time of 18 ms for a batch of 10 images,representing a speed-up of approximately 5.9×10^(4) times compared to exhaustive aperture size enumeration.This work paves the way for the automatic application of aperture photometry in future high-precision surveys such as Tianyu and Legacy Survey of Space and Time.The source code and model are available at https://github.com/ruiyicheng/DeepAP.展开更多
Fast Radio Bursts(FRBs)have emerged as one of the most intriguing and enigmatic phenomena in the field of radio astronomy.The key of current related research is to obtain enough FRB signals.Computer-aided search is ne...Fast Radio Bursts(FRBs)have emerged as one of the most intriguing and enigmatic phenomena in the field of radio astronomy.The key of current related research is to obtain enough FRB signals.Computer-aided search is necessary for that task.Considering the scarcity of FRB signals and massive observation data,the main challenge is about searching speed,accuracy and recall.in this paper,we propose a new FRB search method based on Commensal Radio Astronomy FAST Survey(CRAFTS)data.The CRAFTS drift survey data provide extensive sky coverage and high sensitivity,which significantly enhance the probability of detecting transient signals like FRBs.The search process is separated into two stages on the knowledge of the FRB signal with the structural isomorphism,while a different deep learning model is adopted in each stage.To evaluate the proposed method,FRB signal data sets based on FAST observation data are developed combining simulation FRB signals and real FRB signals.Compared with the benchmark method,the proposed method F-score achieved 0.951,and the associated recall achieved 0.936.The method has been applied to search for FRB signals in raw FAST data.The code and data sets used in the paper are available at github.com/aoxipo.展开更多
In source detection in the Tianlai project,locating the interferometric fringe in visibility data accurately will influence downstream tasks drastically,such as physical parameter estimation and weak source exploratio...In source detection in the Tianlai project,locating the interferometric fringe in visibility data accurately will influence downstream tasks drastically,such as physical parameter estimation and weak source exploration.Considering that traditional locating methods are time-consuming and supervised methods require a great quantity of expensive labeled data,in this paper,we first investigate characteristics of interferometric fringes in the simulation and real scenario separately,and integrate an almost parameter-free unsupervised clustering method and seeding filling or eraser algorithm to propose a hierarchical plug and play method to improve location accuracy.Then,we apply our method to locate single and multiple sources’interferometric fringes in simulation data.Next,we apply our method to real data taken from the Tianlai radio telescope array.Finally,we compare with unsupervised methods that are state of the art.These results show that our method has robustness in different scenarios and can improve location measurement accuracy effectively.展开更多
Radio interferometry significantly improves the resolution of observed images, and the final result also relies heavily on data recovery. The Cotton-Schwab CLEAN(CS-Clean) deconvolution approach is a widely used recon...Radio interferometry significantly improves the resolution of observed images, and the final result also relies heavily on data recovery. The Cotton-Schwab CLEAN(CS-Clean) deconvolution approach is a widely used reconstruction algorithm in the field of radio synthesis imaging. However, parameter tuning for this algorithm has always been a difficult task. Here, its performance is improved by considering some internal characteristics of the data. From a mathematical point of view, a peak signal-to-noise-based(PSNRbased) method was introduced to optimize the step length of the steepest descent method in the recovery process. We also found that the loop gain curve in the new algorithm is a good indicator of parameter tuning.Tests show that the new algorithm can effectively solve the problem of oscillation for a large fixed loop gain and provides a more robust recovery.展开更多
We use the Richardson-Lucy deconvolution algorithm to extract one-dimensional(1 D) spectra from Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST) spectrum images. Compared with other deconvolution algo...We use the Richardson-Lucy deconvolution algorithm to extract one-dimensional(1 D) spectra from Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST) spectrum images. Compared with other deconvolution algorithms, this algorithm is much faster. The application on a real LAMOST image illustrates that the 1 D spectrum resulting from this method has a higher signal-to-noise ratio and resolution than those extracted by the LAMOST pipeline. Furthermore, our algorithm can effectively suppress the ringings that are often present in the 1 D resulting spectra generated by other deconvolution methods.展开更多
基金Astronomy Joint Research Fund supported this work under cooperative agreements between the National Natural Science Foundation of China(NSFC)and the Chinese Academy of Sciences(CAS)(project numbers:U2031132 and U1931206).
文摘This paper proposes a subpixel transformation method to correct Keystone and Smile distortions in fiber spectral images from the Fiber Arrayed Solar Optical Telescope.These distortions affect the spatial and spectral positions,degrading resolution and accuracy.To correct Keystone distortion,we use a local summation and peak-finding method to locate central horizontal coordinates,calculate shifting values,and straighten the curves.For Smile distortion,we use quartic polynomial fitting based on absorption lines at different wavelengths.This technique preserves subpixel components,redistributes pixel values,and interpolates non-fiber portions,rectifying the spectra for accurate analysis.The method can also be applied to other astronomical projects like Large Sky Area Multi-Object Fiber Spectroscopic Telescope,enhancing the accuracy and reliability of spectral data in various astronomical studies.
基金supported by the National SKA Program of China(2020SKA0110300,2020SKA0110201)the National Natural Science Foundation of China(NSFC,grant Nos.12433012 and 12373097)+1 种基金the Guangdong Province Project of the Basic and Applied Basic Research Foundation(2024A1515011503)the Guangzhou Science and Technology Funds(2023A03J0016).
文摘Deconvolution in radio interferometry faces challenges due to incomplete sampling of the visibilities in the spatial frequency domain caused by a limited number of antenna baselines,resulting in an ill-posed inverse problem.Reconstructing dirty images into clean ones is crucial for subsequent scientific analysis.To address these challenges,we propose a U-Net based method that extracts high-level information from the dirty image and reconstructs a clean image by effectively reducing artifacts and sidelobes.The U-Net architecture,consisting of an encoder-decoder structure and skip connections,facilitates the flow of information and preserves spatial details.Using simulated data of radio galaxies,we train our model and evaluate its performance on the testing set.Compared with the CLEAN method and the visibility and image conditioned denoising diffusion probabilistic model,our proposed model can effectively reconstruct both extended sources and faint point sources with higher values in the structural similarity index measure and the peak signal-to-noise ratio.Furthermore,we investigate the impact of noise on the model performance,demonstrating its robustness under varying noise levels.
基金supported by the National Key R&D Program of China(No.2022YFE0133700)the National Natural Science Foundation of China(NSFC,No.12273007)+4 种基金the Guizhou Provincial Excellent Young Science and Technology Talent Program(No.YQK[2023]006)the National SKA Program of China(No.2020SKA0110300)the NSFC(No.11963003)the Guizhou Provincial Basic Research Program(Natural Science)(No.ZK[2022]143)the Cultivation project of Guizhou University(No.[2020]76).
文摘Imaging is an important method for astronomy research.In practice,original images acquired by a telescope are often convolved and blurred by the point-spread function(PSF),which is a very unfavorable situation for many scientific studies including astronomy.This paper introduced a single equation iterative method for solving complex linear equations,and this method can deconvolute dirty images,eliminate the effects of the PSF well.With different PSFs,this algorithm shows very good results in deconvolution.Also,with a giant PSF of aperture synthesis imaging,this algorithm improves the peak signal-to-noise ratio and structural similarity of the dirty images by 41.0%and 33.9%on average.In addition,this paper proves that the algorithm can deconvolute the dirty image by making full use of the information of each pixel in the image,even if the dirty image has salt and pepper noise or even lost areas;by its excellent properties of flexible operation to a single pixel,all these bad situations can be dealt with and the image can be restored.
基金supported by the Youth Program of the Natural Science Foundation of Qinghai Province(2023-ZJ-951Q)Qinghai University Research Ability Enhancement Project(2025KTSQ26).
文摘Aperture photometry is a fundamental technique widely used to obtain high-precision light curves in optical survey projects like Tianyu.However,its effectiveness is limited in crowded fields,and the choice of aperture size critically impacts photometric precision.To address these challenges,we propose DeepAP,an efficient and accurate two-stage deep learning framework for aperture photometry.Specifically,for a given source,we first train a Vision Transformer(ViT)model to assess its feasibility of aperture photometry.We then train the Residual Neural Network(ResNet)to predict its optimal aperture size.For aperture photometry feasibility assessment,the ViT model yields an ROC AUC value of 0.96,and achieves a precision of 0.974,a recall of 0.930,and an F1 score of 0.952 on the test set.For aperture size prediction,the ResNet model effectively mitigates biases inherent in classical growth curve methods by adaptively selecting apertures appropriate for sources of varying brightness,thereby enhancing the signal-to-noise ratio(SNR)across a wide range of targets.Meanwhile,some samples in the test set have a higher SNR than those obtained by exhaustive aperture size enumeration because of the finer granularity of aperture size estimation.By integrating ResNet with the ViT network,the DeepAP framework achieves a median total processing time of 18 ms for a batch of 10 images,representing a speed-up of approximately 5.9×10^(4) times compared to exhaustive aperture size enumeration.This work paves the way for the automatic application of aperture photometry in future high-precision surveys such as Tianyu and Legacy Survey of Space and Time.The source code and model are available at https://github.com/ruiyicheng/DeepAP.
文摘Fast Radio Bursts(FRBs)have emerged as one of the most intriguing and enigmatic phenomena in the field of radio astronomy.The key of current related research is to obtain enough FRB signals.Computer-aided search is necessary for that task.Considering the scarcity of FRB signals and massive observation data,the main challenge is about searching speed,accuracy and recall.in this paper,we propose a new FRB search method based on Commensal Radio Astronomy FAST Survey(CRAFTS)data.The CRAFTS drift survey data provide extensive sky coverage and high sensitivity,which significantly enhance the probability of detecting transient signals like FRBs.The search process is separated into two stages on the knowledge of the FRB signal with the structural isomorphism,while a different deep learning model is adopted in each stage.To evaluate the proposed method,FRB signal data sets based on FAST observation data are developed combining simulation FRB signals and real FRB signals.Compared with the benchmark method,the proposed method F-score achieved 0.951,and the associated recall achieved 0.936.The method has been applied to search for FRB signals in raw FAST data.The code and data sets used in the paper are available at github.com/aoxipo.
基金supported by the National Natural Science Foundation of China(NSFC,grant Nos.42172323 and 12371454)。
文摘In source detection in the Tianlai project,locating the interferometric fringe in visibility data accurately will influence downstream tasks drastically,such as physical parameter estimation and weak source exploration.Considering that traditional locating methods are time-consuming and supervised methods require a great quantity of expensive labeled data,in this paper,we first investigate characteristics of interferometric fringes in the simulation and real scenario separately,and integrate an almost parameter-free unsupervised clustering method and seeding filling or eraser algorithm to propose a hierarchical plug and play method to improve location accuracy.Then,we apply our method to locate single and multiple sources’interferometric fringes in simulation data.Next,we apply our method to real data taken from the Tianlai radio telescope array.Finally,we compare with unsupervised methods that are state of the art.These results show that our method has robustness in different scenarios and can improve location measurement accuracy effectively.
基金partially supported by the Open Research Program of the CAS Key Laboratory of Solar Activity (KLSA201805)the Guizhou Science & Technology Plan Project (Platform Talent No.[2017]5788)+3 种基金the Youth Science & Technology Talents Development Project of Guizhou Education Department (No. KY[2018]119)the National Science Foundation of China (Grant Nos. 11103055, 11773062 and 61605153)“Light of West China” Programme (Grant Nos. RCPY201105 and 2017-XBQNXZ-A-008)the National Basic Research Program of China (973 program: 2012CB821804 and 2015CB857100)
文摘Radio interferometry significantly improves the resolution of observed images, and the final result also relies heavily on data recovery. The Cotton-Schwab CLEAN(CS-Clean) deconvolution approach is a widely used reconstruction algorithm in the field of radio synthesis imaging. However, parameter tuning for this algorithm has always been a difficult task. Here, its performance is improved by considering some internal characteristics of the data. From a mathematical point of view, a peak signal-to-noise-based(PSNRbased) method was introduced to optimize the step length of the steepest descent method in the recovery process. We also found that the loop gain curve in the new algorithm is a good indicator of parameter tuning.Tests show that the new algorithm can effectively solve the problem of oscillation for a large fixed loop gain and provides a more robust recovery.
基金supported by the Joint Research Fund in Astronomy (U1531242) under cooperative agreement between the National Natural Science Foundation of China (NSFC) and Chinese Academy of Sciences (CAS)the NSFC (No. 11673036)+1 种基金the Interdiscipline Research Funds of Beijing Normal UniversityFunding for the project has been provided by the National Development and Reform Commission
文摘We use the Richardson-Lucy deconvolution algorithm to extract one-dimensional(1 D) spectra from Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST) spectrum images. Compared with other deconvolution algorithms, this algorithm is much faster. The application on a real LAMOST image illustrates that the 1 D spectrum resulting from this method has a higher signal-to-noise ratio and resolution than those extracted by the LAMOST pipeline. Furthermore, our algorithm can effectively suppress the ringings that are often present in the 1 D resulting spectra generated by other deconvolution methods.
