Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing imag...Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.展开更多
Based on the analysis of flow characteristics of the FCC riser feedstock injection zone, two novel feedstock injection structures are put forward. By investigating three flow parameters in the feedstock injection zone...Based on the analysis of flow characteristics of the FCC riser feedstock injection zone, two novel feedstock injection structures are put forward. By investigating three flow parameters in the feedstock injection zone under the three different structures (the traditional and the novel No. 1, No. 2 structures): the local density, the particle backmixng ratio, and the jet eigen-concentration, the flow feature under three structures were obtained. The experimental results demonstrate that the flow features under both proposed structures are obviously improved comparing with those under the traditional structure. Especially, the performance of the deflector-structured No. 2 is more desirable than that of No. 1.展开更多
Nowadays, an increasing number of web applications require identification registration. However, the behavior of website registration has not ever been thoroughly studied. We use the database provided by the Chinese S...Nowadays, an increasing number of web applications require identification registration. However, the behavior of website registration has not ever been thoroughly studied. We use the database provided by the Chinese Software Develop Net (CSDN) to provide a complete perspective on this research point. We concentrate on the following three aspects: complexity, correlation, and preference. From these analyses, we draw the following conclusions: firstly, a considerable number of users have not realized the importance of identification and are using very simple identifications that can be attacked very easily. Secondly, there is a strong complexity correlation among the three parts of identification. Thirdly, the top three passwords that users like are 123456789, 12345678 and 11111111, and the top three email providers that they prefer are NETEASE, qq and sina. Further, we provide some suggestions to improve the quality of user passwords.展开更多
Automatic image annotation(AIA)has become an important and challenging problem in computer vision due to the existence of semantic gap.In this paper,a novel support vector machine with mixture of kernels(SVM-MK)for au...Automatic image annotation(AIA)has become an important and challenging problem in computer vision due to the existence of semantic gap.In this paper,a novel support vector machine with mixture of kernels(SVM-MK)for automatic image annotation is proposed.On one hand,the combined global and local block-based image features are extracted in order to reflect the intrinsic content of images as complete as possible.On the other hand,SVM-MK is constructed to shoot for better annotating performance.Experimental results on Corel dataset show that the proposed image feature representation method as well as automatic image annotation classifier,SVM-MK,can achieve higher annotating accuracy than SVM with any single kernel and mi-SVM for semantic image annotation.展开更多
In the present study, we established an UPLC-QTOF-MSE based metabolomic approach in order to evaluate the holistic qualities and compare the quality difference by finding characteristic components of Panax notoginseng...In the present study, we established an UPLC-QTOF-MSE based metabolomic approach in order to evaluate the holistic qualities and compare the quality difference by finding characteristic components of Panax notoginseng extracts (PNE) and Xuesaitong (XST) injection samples from different manufacturers. The data were processed through unsupervised principal component analysis (PCA) and supervised orthogonal partial least squared discrimination analysis (OPLS-DA) to compare the quality differences. Two-dimensional PCA score plots showed a tendency to separate the XST injections and extracts, and most XST injection samples were clearly clustered into two groups. Especially, the injections from He and YB companies were distinguished into two groups. In addition, only injection samples of Hu company were near the cluster of PNE. To explore the potential chemical components contributing most to the differences between XST injection samples from different manufacturers and PNE, an S-plot was constructed following the OPLS-DA. Ginsenoside Rd, ginsenoside Rgl, ginsenoside Re, ginsenoside Rbl, 20(S)-ginsenoside Rhl, gypenoside VII, ginsenoside Rg2, ginsenoside Rh4, ginsenoside Rkl or Rgs, notoginsenoside Fc, 20(R)-ginsenoside Rg3, ginsenoside F2 and protopanaxadiol were recognized as characteristic chemical markers that contributed most to reflect the difference between XST injections and PNE. Ginsenoside Rd, ginsenoside Rgl, ginsenoside Re, ginsenoside Rbl and gypenoside VII were revealed as index components contributing most to the differences of PNE and XST injections, and quantitative analysis of these components could ensure the consistent quality of XST injections. Based on the fact that the injections should be standardized with the characteristic components as quality control chemical markers, it is most important to keep the quality of extracts of raw materials stable and reliable.展开更多
OBJECTIVE: To examine the neuroprotective effect of extract from Naomaitong following focal cerebral ischemia reperfusion induced by occlusion of middle cerebral artery(MCA), and to determine the biochemical alteratio...OBJECTIVE: To examine the neuroprotective effect of extract from Naomaitong following focal cerebral ischemia reperfusion induced by occlusion of middle cerebral artery(MCA), and to determine the biochemical alterations in urine using proton nuclear magnetic resonance spectroscopy and principal component analysis.METHODS: Wistar rats were randomly assigned tothree groups: sham-operated group, MCA focal cerebral ischemia reperfusion model group, and active extract of Naomaitong treatment group. The model was established by an improved MCA occlusion(MCAO) method. Sham-operated rats received the same surgical procedure, but without occlusion. The Naomaitong treatment group were treated with active extract from Naomaitong at a dose of3.0 g·kg-·1d-1. Brain tissues and urine samples were collected from all groups for histopathological assessment and proton nuclear magnetic resonance spectroscopy-based metabonomics, respectively.RESULTS: Hematoxylin-eosin and triphenyl tetrazolium chloride staining of brain tissues showed a significant decrease in cerebral infarction area in the Naomaitong group. In model rats, metabonomic analyses showed increased urinary levels of glutamate, taurine, trimetlylamine oxide, betaine, and glycine, and reduced levels of creatinine and creatine.Naomaitong regulated the metabolic changes by acting on multiple metabolic pathways, including glycine metabolism, glutaminolysis, transmethylation metabolism and creatinine metabolism.CONCLUSION: These data demonstrate that extract from Naomaitong is neuroprotective against focal cerebral ischemia induced by MCAO, and can alleviate biochemical changes in urinary metabolism. Metabonomics may be a useful approach for assessing the biochemical mechanisms underlying the neuroprotective actions of extract from Naomaitong.展开更多
基金supported by National Natural Science Foundation of China(No.61864025)2021 Longyuan Youth Innovation and Entrepreneurship Talent(Team),Young Doctoral Fund of Higher Education Institutions of Gansu Province(No.2021QB-49)+4 种基金Employment and Entrepreneurship Improvement Project of University Students of Gansu Province(No.2021-C-123)Intelligent Tunnel Supervision Robot Research Project(China Railway Scientific Research Institute(Scientific Research)(No.2020-KJ016-Z016-A2)Lanzhou Jiaotong University Youth Foundation(No.2015005)Gansu Higher Education Research Project(No.2016A-018)Gansu Dunhuang Cultural Relics Protection Research Center Open Project(No.GDW2021YB15).
文摘Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.
文摘Based on the analysis of flow characteristics of the FCC riser feedstock injection zone, two novel feedstock injection structures are put forward. By investigating three flow parameters in the feedstock injection zone under the three different structures (the traditional and the novel No. 1, No. 2 structures): the local density, the particle backmixng ratio, and the jet eigen-concentration, the flow feature under three structures were obtained. The experimental results demonstrate that the flow features under both proposed structures are obviously improved comparing with those under the traditional structure. Especially, the performance of the deflector-structured No. 2 is more desirable than that of No. 1.
基金supported by the Foundation for Key Program of Ministry of Education, China under Grant No.311007National Science Foundation Project of China under Grants No. 61202079, No.61170225, No.61271199+1 种基金the Fundamental Research Funds for the Central Universities under Grant No.FRF-TP-09-015Athe Fundamental Research Funds in Beijing Jiaotong University under Grant No.W11JB00630
文摘Nowadays, an increasing number of web applications require identification registration. However, the behavior of website registration has not ever been thoroughly studied. We use the database provided by the Chinese Software Develop Net (CSDN) to provide a complete perspective on this research point. We concentrate on the following three aspects: complexity, correlation, and preference. From these analyses, we draw the following conclusions: firstly, a considerable number of users have not realized the importance of identification and are using very simple identifications that can be attacked very easily. Secondly, there is a strong complexity correlation among the three parts of identification. Thirdly, the top three passwords that users like are 123456789, 12345678 and 11111111, and the top three email providers that they prefer are NETEASE, qq and sina. Further, we provide some suggestions to improve the quality of user passwords.
基金Supported by the National Basic Research Priorities Programme(No.2007CB311004)the National Natural Science Foundation of China(No.61035003,60933004,60903141,60970088,61072085)
文摘Automatic image annotation(AIA)has become an important and challenging problem in computer vision due to the existence of semantic gap.In this paper,a novel support vector machine with mixture of kernels(SVM-MK)for automatic image annotation is proposed.On one hand,the combined global and local block-based image features are extracted in order to reflect the intrinsic content of images as complete as possible.On the other hand,SVM-MK is constructed to shoot for better annotating performance.Experimental results on Corel dataset show that the proposed image feature representation method as well as automatic image annotation classifier,SVM-MK,can achieve higher annotating accuracy than SVM with any single kernel and mi-SVM for semantic image annotation.
文摘In the present study, we established an UPLC-QTOF-MSE based metabolomic approach in order to evaluate the holistic qualities and compare the quality difference by finding characteristic components of Panax notoginseng extracts (PNE) and Xuesaitong (XST) injection samples from different manufacturers. The data were processed through unsupervised principal component analysis (PCA) and supervised orthogonal partial least squared discrimination analysis (OPLS-DA) to compare the quality differences. Two-dimensional PCA score plots showed a tendency to separate the XST injections and extracts, and most XST injection samples were clearly clustered into two groups. Especially, the injections from He and YB companies were distinguished into two groups. In addition, only injection samples of Hu company were near the cluster of PNE. To explore the potential chemical components contributing most to the differences between XST injection samples from different manufacturers and PNE, an S-plot was constructed following the OPLS-DA. Ginsenoside Rd, ginsenoside Rgl, ginsenoside Re, ginsenoside Rbl, 20(S)-ginsenoside Rhl, gypenoside VII, ginsenoside Rg2, ginsenoside Rh4, ginsenoside Rkl or Rgs, notoginsenoside Fc, 20(R)-ginsenoside Rg3, ginsenoside F2 and protopanaxadiol were recognized as characteristic chemical markers that contributed most to reflect the difference between XST injections and PNE. Ginsenoside Rd, ginsenoside Rgl, ginsenoside Re, ginsenoside Rbl and gypenoside VII were revealed as index components contributing most to the differences of PNE and XST injections, and quantitative analysis of these components could ensure the consistent quality of XST injections. Based on the fact that the injections should be standardized with the characteristic components as quality control chemical markers, it is most important to keep the quality of extracts of raw materials stable and reliable.
基金Supported by National Natural Science Foundation of China(Study on the Material Basis and the Ratio of the Effective Components of Naodesheng Based on the Combination of Fingerprint and Metabolic Network,No.81274059Study on the Material Basis of Naomaitong in the Treatment of Ischemic Stroke Based on the in vivo Dynamic Effect and Bioinformatics,No.81274060Study on the in vivo Process and Compatibility Rule of Naomaitong Based on the PK-PD of Effective Components and the Multiobjective Optimization,No.81473413)
文摘OBJECTIVE: To examine the neuroprotective effect of extract from Naomaitong following focal cerebral ischemia reperfusion induced by occlusion of middle cerebral artery(MCA), and to determine the biochemical alterations in urine using proton nuclear magnetic resonance spectroscopy and principal component analysis.METHODS: Wistar rats were randomly assigned tothree groups: sham-operated group, MCA focal cerebral ischemia reperfusion model group, and active extract of Naomaitong treatment group. The model was established by an improved MCA occlusion(MCAO) method. Sham-operated rats received the same surgical procedure, but without occlusion. The Naomaitong treatment group were treated with active extract from Naomaitong at a dose of3.0 g·kg-·1d-1. Brain tissues and urine samples were collected from all groups for histopathological assessment and proton nuclear magnetic resonance spectroscopy-based metabonomics, respectively.RESULTS: Hematoxylin-eosin and triphenyl tetrazolium chloride staining of brain tissues showed a significant decrease in cerebral infarction area in the Naomaitong group. In model rats, metabonomic analyses showed increased urinary levels of glutamate, taurine, trimetlylamine oxide, betaine, and glycine, and reduced levels of creatinine and creatine.Naomaitong regulated the metabolic changes by acting on multiple metabolic pathways, including glycine metabolism, glutaminolysis, transmethylation metabolism and creatinine metabolism.CONCLUSION: These data demonstrate that extract from Naomaitong is neuroprotective against focal cerebral ischemia induced by MCAO, and can alleviate biochemical changes in urinary metabolism. Metabonomics may be a useful approach for assessing the biochemical mechanisms underlying the neuroprotective actions of extract from Naomaitong.