An empirical numerical model that includes nuclear absorption, multiple Coulomb scattering and energy loss is presented for the calculation of transmission through thick objects in high energy proton radiography. In t...An empirical numerical model that includes nuclear absorption, multiple Coulomb scattering and energy loss is presented for the calculation of transmission through thick objects in high energy proton radiography. In this numerical model the angular distributions are treated as Gaussians in the laboratory frame. A Monte Carlo program based on the Geant4 toolkit was developed and used for high energy proton radiography experiment simulations and verification of the empirical numerical model. The two models are used to calculate the transmission fraction of carbon and lead step-wedges in proton radiography at 24 GeV/c, and to calculate radial transmission of the French Test Object in proton radiography at 24 GeV/c with different angular cuts. It is shown that the results of the two models agree with each other, and an analysis of the slight differences is given.展开更多
In this paper,we make use of the boosting method to introduce a new learning algorithm for Gaussian Mixture Models (GMMs) called adapted Boosted Mixture Learning (BML). The method possesses the ability to rectify the ...In this paper,we make use of the boosting method to introduce a new learning algorithm for Gaussian Mixture Models (GMMs) called adapted Boosted Mixture Learning (BML). The method possesses the ability to rectify the existing problems in other conventional techniques for estimating the GMM parameters, due in part to a new mixing-up strategy to increase the number of Gaussian components. The discriminative splitting idea is employed for Gaussian mixture densities followed by learning via the introduced method. Then, the GMM classifier was applied to distinguish between healthy infants and those that present a selected set of medical conditions. Each group includes both full-term and premature infants. Cry-pattern for each pathological condition is created by using the adapted BML method and 13-dimensional Mel-Frequency Cepstral Coefficients (MFCCs) feature vector. The test results demonstrate that the introduced method for training GMMs has a better performance than the traditional method based upon random splitting and EM-based re-estimation as a reference system in multi-pathological classification task.展开更多
A novel free form based face cartoon stylization method is presented in this paper. First, a face cartoon library with marked feature points is constructed. And then select the input image as the target image and an a...A novel free form based face cartoon stylization method is presented in this paper. First, a face cartoon library with marked feature points is constructed. And then select the input image as the target image and an appropriate cartoon image from cartoon library as the reference image, apply the deformation between the corresponding feature points of the images to the target image. Finally, we apply an image stylization process to the result image. As an alternative method, we also choose an appropriate cartoon image as the target image and the input image as the reference image to apply the free from deformation. The experimental results show that our method is straightforward and quick with diversified styles, delivering more infection.展开更多
Recently,many audio search sites headed by Google have used audio fingerprinting technology to search for the same audio and protect the music copyright using one part of the audio data.However,if there are fingerprin...Recently,many audio search sites headed by Google have used audio fingerprinting technology to search for the same audio and protect the music copyright using one part of the audio data.However,if there are fingerprints per audio file,then the amount of query data for the audio search increases.In this paper,we propose a novel method that can reduce the number of fingerprints while providing a level of performance similar to that of existing methods.The proposed method uses the difference of Gaussians which is often used in feature extraction during image signal processing.In the experiment,we use the proposed method and dynamic time warping and undertake an experimental search for the same audio with a success rate of 90%.The proposed method,therefore,can be used for an effective audio search.展开更多
Background The accelerator-driven subcritical transmutation system(ADS)is an advanced technology for the harmless disposal of nuclear waste.A theoretical analysis of the ingredients and content of nuclear waste,partic...Background The accelerator-driven subcritical transmutation system(ADS)is an advanced technology for the harmless disposal of nuclear waste.A theoretical analysis of the ingredients and content of nuclear waste,particularly long-lived waste in a pressurized water reactor(PWR),will provide important information for future spent fuel disposal.Purpose The present study is an attempt to investigate the yields of isotopes in the neutron-induced fission process and estimate the content of long-lived ingredients of nuclear waste in a PWR.Method We combined an approximation of the mass distribution of five Gaussians with the most probable charge model(Zp model)to obtain the isotope yields in the^(235)U(n,f)and^(239)Pu(n,f)processes.The potential energy surface based on the concept of a di-nuclear system model was applied to an approximation using five Gaussian functions.A mathematical formula for the neutron spectrum in a PWR was established,and sets of differential equations were solved to calculate the content of long-lived nuclides in a PWR.Results The calculated isotopic fission yields were in good agreement with the experimental data.Except for 238U,the contents of^(239)Pu,^(240)Pu,^(241)Pu,^(242)Pu,^(237)Np,^(235)U,and^(236)U are predominant in the PWR after reaching a discharge burnup.In addition,some isotope pairs of heavy nuclei reach a similar value after stabilization,which can be explained by the decay chain and effective fission crosssections.For fission fragments,we simulated the content evolution of some long-lived nuclides^(90)Sr,^(107)Pd,^(135)Cs,and their isobars^(90)Rb,^(107)Rh,and^(135)Xe during a fuel cycle in a PWR.The variations in the inventories of uranium and plutonium were in good agreement with the data in Daya Bay.Conclusion A new method is proposed for the prediction of the isotopic fission yield.The inventory of long-lived nuclides was analyzed and predicted after reaching a discharge burnup.展开更多
Due to the high accuracy and fast acquisition speed offered by airborne Light Detection and Ranging(LiDAR)technology,airborne LiDAR point clouds have been widely used in three-dimensional building model reconstruction...Due to the high accuracy and fast acquisition speed offered by airborne Light Detection and Ranging(LiDAR)technology,airborne LiDAR point clouds have been widely used in three-dimensional building model reconstruction.This paper presents a novel approach to segment building roofs from point clouds using a Gaussian mixture model in which buildings are represented by a mixture of Gaussians(MoG).The Expectation-Maximization(EM)algorithm with the minimum description length(MDL)principle is employed to obtain the optimal parameters of the MoG model for separating building roofs.To separate complete planar building roofs,coplanar Gaussian components are merged according to their distances to the corresponding planes.In addition,shape analysis is utilized to remove nonplanar objects caused by trees and irregular artifacts.Building models are obtained by combining segmented planar roofs,topological relationships,and regularized building boundaries.Roof intersection segments and points are derived by the segmentation results,and a rasterbased regularization method is employed to obtain geometrically correct and regular building models.Experimental results suggest that the segmentation method is able to separate building roofs with high accuracy while maintaining correct topological relationships among roofs.展开更多
高斯混合模型广泛应用于基于背景建模的运动目标检测中,高斯混合模型参数估计和更新算法影响到背景模型的性能。文中对传统的高斯混合背景模型进行了改进,针对背景局部运动、活动阴影等问题,采用混合色彩值抑制阴影,在背景更新中引入一...高斯混合模型广泛应用于基于背景建模的运动目标检测中,高斯混合模型参数估计和更新算法影响到背景模型的性能。文中对传统的高斯混合背景模型进行了改进,针对背景局部运动、活动阴影等问题,采用混合色彩值抑制阴影,在背景更新中引入一个"前景支撑映射"(Foreground Support Map,FSB),较好地解决了背景模型的提取、更新、背景扰动、外界光照变化等问题。实验结果证明,实验结果验证了该方法的有效性和在复杂背景变化下的鲁棒性。展开更多
基金Supported by NSAF(11176001)Science and Technology Developing Foundation of China Academy of Engineering Physics(2012A0202006)
文摘An empirical numerical model that includes nuclear absorption, multiple Coulomb scattering and energy loss is presented for the calculation of transmission through thick objects in high energy proton radiography. In this numerical model the angular distributions are treated as Gaussians in the laboratory frame. A Monte Carlo program based on the Geant4 toolkit was developed and used for high energy proton radiography experiment simulations and verification of the empirical numerical model. The two models are used to calculate the transmission fraction of carbon and lead step-wedges in proton radiography at 24 GeV/c, and to calculate radial transmission of the French Test Object in proton radiography at 24 GeV/c with different angular cuts. It is shown that the results of the two models agree with each other, and an analysis of the slight differences is given.
文摘In this paper,we make use of the boosting method to introduce a new learning algorithm for Gaussian Mixture Models (GMMs) called adapted Boosted Mixture Learning (BML). The method possesses the ability to rectify the existing problems in other conventional techniques for estimating the GMM parameters, due in part to a new mixing-up strategy to increase the number of Gaussian components. The discriminative splitting idea is employed for Gaussian mixture densities followed by learning via the introduced method. Then, the GMM classifier was applied to distinguish between healthy infants and those that present a selected set of medical conditions. Each group includes both full-term and premature infants. Cry-pattern for each pathological condition is created by using the adapted BML method and 13-dimensional Mel-Frequency Cepstral Coefficients (MFCCs) feature vector. The test results demonstrate that the introduced method for training GMMs has a better performance than the traditional method based upon random splitting and EM-based re-estimation as a reference system in multi-pathological classification task.
文摘A novel free form based face cartoon stylization method is presented in this paper. First, a face cartoon library with marked feature points is constructed. And then select the input image as the target image and an appropriate cartoon image from cartoon library as the reference image, apply the deformation between the corresponding feature points of the images to the target image. Finally, we apply an image stylization process to the result image. As an alternative method, we also choose an appropriate cartoon image as the target image and the input image as the reference image to apply the free from deformation. The experimental results show that our method is straightforward and quick with diversified styles, delivering more infection.
文摘Recently,many audio search sites headed by Google have used audio fingerprinting technology to search for the same audio and protect the music copyright using one part of the audio data.However,if there are fingerprints per audio file,then the amount of query data for the audio search increases.In this paper,we propose a novel method that can reduce the number of fingerprints while providing a level of performance similar to that of existing methods.The proposed method uses the difference of Gaussians which is often used in feature extraction during image signal processing.In the experiment,we use the proposed method and dynamic time warping and undertake an experimental search for the same audio with a success rate of 90%.The proposed method,therefore,can be used for an effective audio search.
基金This work was supported by the National Natural Science Foundation of China(No.11605296)the Natural Science Foundation of Guangdong Province,China(No.2016A030310208)Fundamental Research Funds for Central Universities(No.18lgpy87).
文摘Background The accelerator-driven subcritical transmutation system(ADS)is an advanced technology for the harmless disposal of nuclear waste.A theoretical analysis of the ingredients and content of nuclear waste,particularly long-lived waste in a pressurized water reactor(PWR),will provide important information for future spent fuel disposal.Purpose The present study is an attempt to investigate the yields of isotopes in the neutron-induced fission process and estimate the content of long-lived ingredients of nuclear waste in a PWR.Method We combined an approximation of the mass distribution of five Gaussians with the most probable charge model(Zp model)to obtain the isotope yields in the^(235)U(n,f)and^(239)Pu(n,f)processes.The potential energy surface based on the concept of a di-nuclear system model was applied to an approximation using five Gaussian functions.A mathematical formula for the neutron spectrum in a PWR was established,and sets of differential equations were solved to calculate the content of long-lived nuclides in a PWR.Results The calculated isotopic fission yields were in good agreement with the experimental data.Except for 238U,the contents of^(239)Pu,^(240)Pu,^(241)Pu,^(242)Pu,^(237)Np,^(235)U,and^(236)U are predominant in the PWR after reaching a discharge burnup.In addition,some isotope pairs of heavy nuclei reach a similar value after stabilization,which can be explained by the decay chain and effective fission crosssections.For fission fragments,we simulated the content evolution of some long-lived nuclides^(90)Sr,^(107)Pd,^(135)Cs,and their isobars^(90)Rb,^(107)Rh,and^(135)Xe during a fuel cycle in a PWR.The variations in the inventories of uranium and plutonium were in good agreement with the data in Daya Bay.Conclusion A new method is proposed for the prediction of the isotopic fission yield.The inventory of long-lived nuclides was analyzed and predicted after reaching a discharge burnup.
基金This work was supported by the National Key Scientific Instrument and Equipment Development Projects of China[grant number 2013YQ120343]the National Natural Science Foundation of China[grant numbers 41171265 and 41101436]the 100 Talents Program of the Chinese Academy of Sciences.
文摘Due to the high accuracy and fast acquisition speed offered by airborne Light Detection and Ranging(LiDAR)technology,airborne LiDAR point clouds have been widely used in three-dimensional building model reconstruction.This paper presents a novel approach to segment building roofs from point clouds using a Gaussian mixture model in which buildings are represented by a mixture of Gaussians(MoG).The Expectation-Maximization(EM)algorithm with the minimum description length(MDL)principle is employed to obtain the optimal parameters of the MoG model for separating building roofs.To separate complete planar building roofs,coplanar Gaussian components are merged according to their distances to the corresponding planes.In addition,shape analysis is utilized to remove nonplanar objects caused by trees and irregular artifacts.Building models are obtained by combining segmented planar roofs,topological relationships,and regularized building boundaries.Roof intersection segments and points are derived by the segmentation results,and a rasterbased regularization method is employed to obtain geometrically correct and regular building models.Experimental results suggest that the segmentation method is able to separate building roofs with high accuracy while maintaining correct topological relationships among roofs.
文摘高斯混合模型广泛应用于基于背景建模的运动目标检测中,高斯混合模型参数估计和更新算法影响到背景模型的性能。文中对传统的高斯混合背景模型进行了改进,针对背景局部运动、活动阴影等问题,采用混合色彩值抑制阴影,在背景更新中引入一个"前景支撑映射"(Foreground Support Map,FSB),较好地解决了背景模型的提取、更新、背景扰动、外界光照变化等问题。实验结果证明,实验结果验证了该方法的有效性和在复杂背景变化下的鲁棒性。
