AIM:To establish a recording system with a direct view of the surgeon to supplement video recording under an operating microscope,which lacks information on the movement and position of the surgeon’s hands,and to fac...AIM:To establish a recording system with a direct view of the surgeon to supplement video recording under an operating microscope,which lacks information on the movement and position of the surgeon’s hands,and to facilitate the reproduction of a skilled surgeon’s technique by a surgeon in training.METHODS:A small camera was attached to the operating microscope with a custom adapter.Microscopic surgeon’s view and direct surgeon’s view through this new camera were recorded in the surgical recording system.Both movies were synchronized and analyzed how do surgeons handle the instruments.RESULTS:A small camera attached to the operating microscope allowed the surgeon’s hands motion to be recorded without interfering with the surgeon’s movements.Different surgeons used different methods to manipulate the ultrasound handpiece and the irrigation/aspiration device.Even in the simple paracentesis procedure,different surgeons used different methods.Surgeons-in-training were able to identify and improve their weaknesses by watching synchronized movies of their hand motions and microscopic view.CONCLUSION:Simultaneous recording the surgical field out of the operating microscopic view by a small camera set on the microscope is comprehensive and improves surgeons-in-training understanding and learning surgeries.展开更多
为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度...为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度特征提取模块,获取到具有全局上下文信息和位置信息的多尺度特征图像。通过残差学习网络对深度图进行优化,防止多次卷积操作出现重建边缘模糊的问题。通过分类的思想构建focal loss函数增强网络模型的判断能力。由实验结果可知,该算法在DTU(technical university of denmark)数据集上和CasMVSNet(Cascade MVSNet)算法相比,在整体精度误差、运行时间、显存资源占用上分别降低了14.08%、72.15%、4.62%。在Tanks and Temples数据集整体评价指标Mean上该模型优于其他算法,证明提出的基于自适应空间特征增强的多视图深度估计算法的有效性。展开更多
In this paper a millimeter-wave （MMW） squint indirect holographic method is presented, which is suitable for imaging with a large field-of-view. The proposed system employs the squint operation mode to remove the ba...In this paper a millimeter-wave （MMW） squint indirect holographic method is presented, which is suitable for imaging with a large field-of-view. The proposed system employs the squint operation mode to remove the background and twin- image interferences, which achieves a similar effect to off-axis holography but leaves out the large-aperture quasi-optical component. The translational scanning manner enables a large field of view and ensures the image uniformity, which is difficult to realize in off-axis holography. In addition, a corresponding imaging algorithm for the presented scheme is developed to reconstruct the image from the recorded hologram. Some imaging results on typical objects, obtained with electromagnetic simulation, demonstrate good performance of the imaging scheme and validate the effectiveness of the image reconstruction algorithm.展开更多
动车组运用检修物料清单(BOM,Bill of Material)是动车组运用检修信息化管理的基础,是开展动车组健康管理的前提。文章在研究动车组运用检修BOM应用需求的基础上,提出了一种动车组运用检修BOM的设计、编制和变更管理方法,介绍了该BOM在...动车组运用检修物料清单(BOM,Bill of Material)是动车组运用检修信息化管理的基础,是开展动车组健康管理的前提。文章在研究动车组运用检修BOM应用需求的基础上,提出了一种动车组运用检修BOM的设计、编制和变更管理方法,介绍了该BOM在辅助检修作业、数据串联、信息采集等动车组运用检修管理业务中的应用。为开展全谱系动车组运用检修BOM编制工作提供了参考。展开更多
Flume, which implements decentralized information flow control （DIFC）, allows a high security level process to ＂pre-create＂ secret files in a low security level directory. However, the pre-create mechanism makes s...Flume, which implements decentralized information flow control （DIFC）, allows a high security level process to ＂pre-create＂ secret files in a low security level directory. However, the pre-create mechanism makes some normal system calls unavailable, and moreover, it needs priori knowledge to create a large quantity of objects, which is difficult to estimate in practical operating systems. In this paper, we present an extended Flume file access control mechanism, named Effect, to substitute the mechanism of pre-create, which permits write operations （create, delete, and rename a file） on directories and creates a file access virtual layer that allocates operational views for each process with noninterference properties. In the end, we further present an analysis on the security of Effect. Our work makes it easier for multi-user to share confidential information in decentralized information flow control systems.展开更多
文摘AIM:To establish a recording system with a direct view of the surgeon to supplement video recording under an operating microscope,which lacks information on the movement and position of the surgeon’s hands,and to facilitate the reproduction of a skilled surgeon’s technique by a surgeon in training.METHODS:A small camera was attached to the operating microscope with a custom adapter.Microscopic surgeon’s view and direct surgeon’s view through this new camera were recorded in the surgical recording system.Both movies were synchronized and analyzed how do surgeons handle the instruments.RESULTS:A small camera attached to the operating microscope allowed the surgeon’s hands motion to be recorded without interfering with the surgeon’s movements.Different surgeons used different methods to manipulate the ultrasound handpiece and the irrigation/aspiration device.Even in the simple paracentesis procedure,different surgeons used different methods.Surgeons-in-training were able to identify and improve their weaknesses by watching synchronized movies of their hand motions and microscopic view.CONCLUSION:Simultaneous recording the surgical field out of the operating microscopic view by a small camera set on the microscope is comprehensive and improves surgeons-in-training understanding and learning surgeries.
文摘为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度特征提取模块,获取到具有全局上下文信息和位置信息的多尺度特征图像。通过残差学习网络对深度图进行优化,防止多次卷积操作出现重建边缘模糊的问题。通过分类的思想构建focal loss函数增强网络模型的判断能力。由实验结果可知,该算法在DTU(technical university of denmark)数据集上和CasMVSNet(Cascade MVSNet)算法相比,在整体精度误差、运行时间、显存资源占用上分别降低了14.08%、72.15%、4.62%。在Tanks and Temples数据集整体评价指标Mean上该模型优于其他算法,证明提出的基于自适应空间特征增强的多视图深度估计算法的有效性。
基金Project supported by the National Natural Science Foundation of China (Grant Nos.11174280,60990323,and 60990320)the Knowledge Innovation Program of the Chinese Academy of Sciences (Grant No.YYYJ-1123)
文摘In this paper a millimeter-wave （MMW） squint indirect holographic method is presented, which is suitable for imaging with a large field-of-view. The proposed system employs the squint operation mode to remove the background and twin- image interferences, which achieves a similar effect to off-axis holography but leaves out the large-aperture quasi-optical component. The translational scanning manner enables a large field of view and ensures the image uniformity, which is difficult to realize in off-axis holography. In addition, a corresponding imaging algorithm for the presented scheme is developed to reconstruct the image from the recorded hologram. Some imaging results on typical objects, obtained with electromagnetic simulation, demonstrate good performance of the imaging scheme and validate the effectiveness of the image reconstruction algorithm.
文摘动车组运用检修物料清单(BOM,Bill of Material)是动车组运用检修信息化管理的基础,是开展动车组健康管理的前提。文章在研究动车组运用检修BOM应用需求的基础上,提出了一种动车组运用检修BOM的设计、编制和变更管理方法,介绍了该BOM在辅助检修作业、数据串联、信息采集等动车组运用检修管理业务中的应用。为开展全谱系动车组运用检修BOM编制工作提供了参考。
基金Supported by the National Natural Science Foundation of China(61003268,61103220,91118003,61173138,61170022)Hubei Provincial Natural Science Foundation(2010CDB08601)The Fundamental ResearchFunds for the Central Universities (3101038,274629)
文摘Flume, which implements decentralized information flow control （DIFC）, allows a high security level process to ＂pre-create＂ secret files in a low security level directory. However, the pre-create mechanism makes some normal system calls unavailable, and moreover, it needs priori knowledge to create a large quantity of objects, which is difficult to estimate in practical operating systems. In this paper, we present an extended Flume file access control mechanism, named Effect, to substitute the mechanism of pre-create, which permits write operations （create, delete, and rename a file） on directories and creates a file access virtual layer that allocates operational views for each process with noninterference properties. In the end, we further present an analysis on the security of Effect. Our work makes it easier for multi-user to share confidential information in decentralized information flow control systems.
