The Five-hundred-meter Aperture Spherical radio Telescope(FAST)is the world’s largest single-dish radio telescope.Its large reflecting surface achieves unprecedented sensitivity but is prone to damage,such as dents a...The Five-hundred-meter Aperture Spherical radio Telescope(FAST)is the world’s largest single-dish radio telescope.Its large reflecting surface achieves unprecedented sensitivity but is prone to damage,such as dents and holes,caused by naturally-occurring falling objects.Hence,the timely and accurate detection of surface defects is crucial for FAST’s stable operation.Conventional manual inspection involves human inspectors climbing up and examining the large surface visually,a time-consuming and potentially unreliable process.To accelerate the inspection process and increase its accuracy,this work makes the first step towards automating the inspection of FAST by integrating deep-learning techniques with drone technology.First,a drone flies over the surface along a predetermined route.Since surface defects significantly vary in scale and show high inter-class similarity,directly applying existing deep detectors to detect defects on the drone imagery is highly prone to missing and misidentifying defects.As a remedy,we introduce cross-fusion,a dedicated plug-in operation for deep detectors that enables the adaptive fusion of multi-level features in a point-wise selective fashion,depending on local defect patterns.Consequently,strong semantics and fine-grained details are dynamically fused at different positions to support the accurate detection of defects of various scales and types.Our AI-powered drone-based automated inspection is time-efficient,reliable,and has good accessibility,which guarantees the long-term and stable operation of FAST.展开更多
In contrast to ion beams produced by conventional accelerators,ion beams accelerated by ultrashort intense laser pulses have advantages of ultrashort bunch duration and ultrahigh density,which are achieved in compact ...In contrast to ion beams produced by conventional accelerators,ion beams accelerated by ultrashort intense laser pulses have advantages of ultrashort bunch duration and ultrahigh density,which are achieved in compact size.However,it is still challenging to simultaneously enhance their quality and yield for practical applications such as fast ion ignition of inertial confinement fusion.Compared with other mechanisms of laser-driven ion acceleration,the hole-boring radiation pressure acceleration has a special advantage in generating high-fluence ion beams suitable for the creation of high energy density state of matters.In this paper,we present a review on some theoretical and numerical studies of the hole-boring radiation pressure acceleration.First we discuss the typical field structure associated with this mechanism,its intrinsic feature of oscillations,and the underling physics.Then we will review some recently proposed schemes to enhance the beam quality and the efficiency in the hole-boring radiation pressure acceleration,such as matching laser intensity profile with target density profile,and using two-ion-species targets.Based on this,we propose an integrated scheme for efficient high-quality hole-boring radiation pressure acceleration,in which the longitudinal density profile of a composite target as well as the laser transverse intensity profile are tailored according to the matching condition.展开更多
An algorithm based on deep semantic segmentation called LC-DeepLab is proposed for detecting the trends and geometries of cracks on tunnel linings at the pixel level.The proposed method addresses the low accuracy of t...An algorithm based on deep semantic segmentation called LC-DeepLab is proposed for detecting the trends and geometries of cracks on tunnel linings at the pixel level.The proposed method addresses the low accuracy of tunnel crack segmentation and the slow detection speed of conventional models in complex backgrounds.The novel algorithm is based on the DeepLabv3+network framework.A lighter backbone network was used for feature extraction.Next,an efficient shallow feature fusion module that extracts crack features across pixels is designed to improve the edges of crack segmentation.Finally,an efficient attention module that significantly improves the anti-interference ability of the model in complex backgrounds is validated.Four classic semantic segmentation algorithms(fully convolutional network,pyramid scene parsing network,U-Net,and DeepLabv3+)are selected for comparative analysis to verify the effectiveness of the proposed algorithm.The experimental results show that LC-DeepLab can accurately segment and highlight cracks from tunnel linings in complex backgrounds,and the accuracy(mean intersection over union)is 78.26%.The LC-DeepLab can achieve a real-time segmentation of 416×416×3 defect images with 46.98 f/s and 21.85 Mb parameters.展开更多
A novel multi-focus polychromatic image fusion algorithm based on filtering in the frequency domain using fast Fourier transform(FFT) and synthesis in the space domain(FFDSSD) is presented in this paper.First,the orig...A novel multi-focus polychromatic image fusion algorithm based on filtering in the frequency domain using fast Fourier transform(FFT) and synthesis in the space domain(FFDSSD) is presented in this paper.First,the original multi-focus images are transformed into their frequency data by FFT for easy and accurate clarity determination.Then a Gaussian low-pass filter is used to filter the high frequency information corresponding to the image saliencies.After an inverse FFT,the filtered images are obtained.The deviation between the filtered images and the original ones,representing the clarity of the image,is used to select the pixels from the multi-focus images to reconstruct a completely focused image.These operations in space domain preserve the original information as much as possible and are relatively insensitive to misregistration scenarios with respect to transform domain methods.The polychromatic noise is well considered and successfully avoided while the information in different chromatic channels is preserved.A natural,nice-looking fused microscopic image for human visual evaluations is obtained in a dedicated experiment.The experimental results indicate that the proposed algorithm has a good performance in objective quality metrics and runtime efficiency.展开更多
We discuss a hardship in synthesis of heaviest super heavy elements in massive nuclei reactions due to the hindrance to complete fusion of reacting nuclei caused on the onset of quasifission process which strongly com...We discuss a hardship in synthesis of heaviest super heavy elements in massive nuclei reactions due to the hindrance to complete fusion of reacting nuclei caused on the onset of quasifission process which strongly competes with complete fusion and due to the strong increase of fission yields along the de-excitation cascade of the compound nucleus in comparison with the evaporation residue formation.The hindrance to formation of compound nucleus and evaporation residue is determined by the characteristic of the entrance channel.展开更多
基金financially supported by the National Natural Science Foundation of China(No.62101032)the Postdoctoral Science Foundation of China(Nos.2021M690015,2022T150050)the Beijing Institute of Technology Research Fund Program for Young Scholars(No.3040011182111).
文摘The Five-hundred-meter Aperture Spherical radio Telescope(FAST)is the world’s largest single-dish radio telescope.Its large reflecting surface achieves unprecedented sensitivity but is prone to damage,such as dents and holes,caused by naturally-occurring falling objects.Hence,the timely and accurate detection of surface defects is crucial for FAST’s stable operation.Conventional manual inspection involves human inspectors climbing up and examining the large surface visually,a time-consuming and potentially unreliable process.To accelerate the inspection process and increase its accuracy,this work makes the first step towards automating the inspection of FAST by integrating deep-learning techniques with drone technology.First,a drone flies over the surface along a predetermined route.Since surface defects significantly vary in scale and show high inter-class similarity,directly applying existing deep detectors to detect defects on the drone imagery is highly prone to missing and misidentifying defects.As a remedy,we introduce cross-fusion,a dedicated plug-in operation for deep detectors that enables the adaptive fusion of multi-level features in a point-wise selective fashion,depending on local defect patterns.Consequently,strong semantics and fine-grained details are dynamically fused at different positions to support the accurate detection of defects of various scales and types.Our AI-powered drone-based automated inspection is time-efficient,reliable,and has good accessibility,which guarantees the long-term and stable operation of FAST.
基金This work was supported in part by the National Basic Research Program of China(Grant No.2013CBA01504)the National Natural Science Foundation of China(Grant Nos.11675108,11421064,11405108 and 11374210).
文摘In contrast to ion beams produced by conventional accelerators,ion beams accelerated by ultrashort intense laser pulses have advantages of ultrashort bunch duration and ultrahigh density,which are achieved in compact size.However,it is still challenging to simultaneously enhance their quality and yield for practical applications such as fast ion ignition of inertial confinement fusion.Compared with other mechanisms of laser-driven ion acceleration,the hole-boring radiation pressure acceleration has a special advantage in generating high-fluence ion beams suitable for the creation of high energy density state of matters.In this paper,we present a review on some theoretical and numerical studies of the hole-boring radiation pressure acceleration.First we discuss the typical field structure associated with this mechanism,its intrinsic feature of oscillations,and the underling physics.Then we will review some recently proposed schemes to enhance the beam quality and the efficiency in the hole-boring radiation pressure acceleration,such as matching laser intensity profile with target density profile,and using two-ion-species targets.Based on this,we propose an integrated scheme for efficient high-quality hole-boring radiation pressure acceleration,in which the longitudinal density profile of a composite target as well as the laser transverse intensity profile are tailored according to the matching condition.
基金This study was supported by the National Natural Science Foundation of China(Grant Nos.50908234,52208421)the Open Fund of the National Engineering Research Center of Highway Maintenance Technology,Changsha University of Science&Technology(No.kfj220101)+1 种基金the Natural Science Foundation of Hunan Province(No.2020JJ4743)the Research Innovation Project for Postgraduate of Central South University(No.1053320213484).
文摘An algorithm based on deep semantic segmentation called LC-DeepLab is proposed for detecting the trends and geometries of cracks on tunnel linings at the pixel level.The proposed method addresses the low accuracy of tunnel crack segmentation and the slow detection speed of conventional models in complex backgrounds.The novel algorithm is based on the DeepLabv3+network framework.A lighter backbone network was used for feature extraction.Next,an efficient shallow feature fusion module that extracts crack features across pixels is designed to improve the edges of crack segmentation.Finally,an efficient attention module that significantly improves the anti-interference ability of the model in complex backgrounds is validated.Four classic semantic segmentation algorithms(fully convolutional network,pyramid scene parsing network,U-Net,and DeepLabv3+)are selected for comparative analysis to verify the effectiveness of the proposed algorithm.The experimental results show that LC-DeepLab can accurately segment and highlight cracks from tunnel linings in complex backgrounds,and the accuracy(mean intersection over union)is 78.26%.The LC-DeepLab can achieve a real-time segmentation of 416×416×3 defect images with 46.98 f/s and 21.85 Mb parameters.
文摘A novel multi-focus polychromatic image fusion algorithm based on filtering in the frequency domain using fast Fourier transform(FFT) and synthesis in the space domain(FFDSSD) is presented in this paper.First,the original multi-focus images are transformed into their frequency data by FFT for easy and accurate clarity determination.Then a Gaussian low-pass filter is used to filter the high frequency information corresponding to the image saliencies.After an inverse FFT,the filtered images are obtained.The deviation between the filtered images and the original ones,representing the clarity of the image,is used to select the pixels from the multi-focus images to reconstruct a completely focused image.These operations in space domain preserve the original information as much as possible and are relatively insensitive to misregistration scenarios with respect to transform domain methods.The polychromatic noise is well considered and successfully avoided while the information in different chromatic channels is preserved.A natural,nice-looking fused microscopic image for human visual evaluations is obtained in a dedicated experiment.The experimental results indicate that the proposed algorithm has a good performance in objective quality metrics and runtime efficiency.
文摘We discuss a hardship in synthesis of heaviest super heavy elements in massive nuclei reactions due to the hindrance to complete fusion of reacting nuclei caused on the onset of quasifission process which strongly competes with complete fusion and due to the strong increase of fission yields along the de-excitation cascade of the compound nucleus in comparison with the evaporation residue formation.The hindrance to formation of compound nucleus and evaporation residue is determined by the characteristic of the entrance channel.
