为了提升导弹在复杂环境下的寻的制导能力,设计了一种红外与可见光双模式导引头光学系统。该方案中采用分光镜透射红外光反射可见光,使结构布局更加紧凑,实现红外与可见光共口径,同时配合红外材料选取,实现光学被动消热差设计。中红外...为了提升导弹在复杂环境下的寻的制导能力,设计了一种红外与可见光双模式导引头光学系统。该方案中采用分光镜透射红外光反射可见光,使结构布局更加紧凑,实现红外与可见光共口径,同时配合红外材料选取,实现光学被动消热差设计。中红外模式视场角3°×2.3°,可见光模式视场角5°×4°,工作温度20℃条件下,双模式在截止频率处,MTF(Modulation Transfer Function)值均大于0.4。红外与可见光双模式光学系统适合应用于复杂环境的导弹制导,对温度有良好的适应性,具有较好的成像质量,满足系统的性能要求。展开更多
This research article introduces and explores the concept of a hybrid prism, which combines the properties of a lens and a reflective prism, designed for optical systems that operate in different spectral ranges of el...This research article introduces and explores the concept of a hybrid prism, which combines the properties of a lens and a reflective prism, designed for optical systems that operate in different spectral ranges of electromagnetic waves. The hybrid prism allows for precise focusing of light rays in a glass body and X-rays in a vacuum, enabling it to serve as an objective in various optical systems for imaging objects. The article delves into the structure and working principles of the hybrid prism, discussing its potential applications, including as an intraocular prism for macular degeneration, a lidar system for vehicle navigation, and objectives for cameras, telescopes, microscopes, X-ray devices, and X-ray microscopes. The revolutionary hybrid prism unlocks precise imaging of light and X-rays, reshaping optical systems and enabling groundbreaking applications.展开更多
近些年来,在目标检测以及图像分割等领域涌现了许多先进的算法。在能见度较差的微光场景下如夜晚、大雾天气等场景中,视频图像具有像素声高、对比度低、无彩色信息等特点,算法的检测性能受到明显限制。与目前主流的RGB相机相比,毫米波...近些年来,在目标检测以及图像分割等领域涌现了许多先进的算法。在能见度较差的微光场景下如夜晚、大雾天气等场景中,视频图像具有像素声高、对比度低、无彩色信息等特点,算法的检测性能受到明显限制。与目前主流的RGB相机相比,毫米波雷达对上述复杂环境具有一定的免疫能力,可以在不利条件下辅助RGB相机进行目标检测工作。以单阶段目标检测器中实时性较高的YOLOv5s为基础,结合毫米波雷达的特性,提出了用于微光环境下目标检测的多模态识别网络。与现有的传感器融合方法相比,多模态识别网络有几个关键优势。系统以基于学习的方式融合了2种模态,只需要少量新场景的标记图像和雷达数据,因为其可以充分利用已经开源的大型图像数据集进行大批量的训练。这一突出特性使新系统能够适应高度复杂的现实环境。由于使用了高度计算效率的融合方法,系统是非常轻量级的,因此适用于各个复杂场景下的实时应用。为了评估系统的性能,制作了一个小批量的雷达和摄像机融合数据集,包含普通光照和不同强度微光光照条件下的多模态数据。实验结果表明,微光场景下多模态识别网络的平均精度达到76.6%,相比Faster R-CNN算法和YOLOv7算法,全类平均精度(mean Average Precision,mAP)提高了16.8%和9.3%,且误检、漏检率低,达到了在微光环境下完成目标检测任务的要求。展开更多
The wave equation of the electron, recently improved, allows physics to obtain all the quantum numbers and other results explaining the hydrogen spectrum. The Pauli exclusion principle then gives the description of el...The wave equation of the electron, recently improved, allows physics to obtain all the quantum numbers and other results explaining the hydrogen spectrum. The Pauli exclusion principle then gives the description of electron clouds used in chemistry. The relativistic wave equation is associated with a Lagrangian density, thus also with an energy-momentum tensorial density. The wave of an electron cloud adds these energy-momentum densities, while photons in light are precisely those differences between such energy-momentum densities.展开更多
文摘为了提升导弹在复杂环境下的寻的制导能力,设计了一种红外与可见光双模式导引头光学系统。该方案中采用分光镜透射红外光反射可见光,使结构布局更加紧凑,实现红外与可见光共口径,同时配合红外材料选取,实现光学被动消热差设计。中红外模式视场角3°×2.3°,可见光模式视场角5°×4°,工作温度20℃条件下,双模式在截止频率处,MTF(Modulation Transfer Function)值均大于0.4。红外与可见光双模式光学系统适合应用于复杂环境的导弹制导,对温度有良好的适应性,具有较好的成像质量,满足系统的性能要求。
文摘This research article introduces and explores the concept of a hybrid prism, which combines the properties of a lens and a reflective prism, designed for optical systems that operate in different spectral ranges of electromagnetic waves. The hybrid prism allows for precise focusing of light rays in a glass body and X-rays in a vacuum, enabling it to serve as an objective in various optical systems for imaging objects. The article delves into the structure and working principles of the hybrid prism, discussing its potential applications, including as an intraocular prism for macular degeneration, a lidar system for vehicle navigation, and objectives for cameras, telescopes, microscopes, X-ray devices, and X-ray microscopes. The revolutionary hybrid prism unlocks precise imaging of light and X-rays, reshaping optical systems and enabling groundbreaking applications.
文摘近些年来,在目标检测以及图像分割等领域涌现了许多先进的算法。在能见度较差的微光场景下如夜晚、大雾天气等场景中,视频图像具有像素声高、对比度低、无彩色信息等特点,算法的检测性能受到明显限制。与目前主流的RGB相机相比,毫米波雷达对上述复杂环境具有一定的免疫能力,可以在不利条件下辅助RGB相机进行目标检测工作。以单阶段目标检测器中实时性较高的YOLOv5s为基础,结合毫米波雷达的特性,提出了用于微光环境下目标检测的多模态识别网络。与现有的传感器融合方法相比,多模态识别网络有几个关键优势。系统以基于学习的方式融合了2种模态,只需要少量新场景的标记图像和雷达数据,因为其可以充分利用已经开源的大型图像数据集进行大批量的训练。这一突出特性使新系统能够适应高度复杂的现实环境。由于使用了高度计算效率的融合方法,系统是非常轻量级的,因此适用于各个复杂场景下的实时应用。为了评估系统的性能,制作了一个小批量的雷达和摄像机融合数据集,包含普通光照和不同强度微光光照条件下的多模态数据。实验结果表明,微光场景下多模态识别网络的平均精度达到76.6%,相比Faster R-CNN算法和YOLOv7算法,全类平均精度(mean Average Precision,mAP)提高了16.8%和9.3%,且误检、漏检率低,达到了在微光环境下完成目标检测任务的要求。
文摘The wave equation of the electron, recently improved, allows physics to obtain all the quantum numbers and other results explaining the hydrogen spectrum. The Pauli exclusion principle then gives the description of electron clouds used in chemistry. The relativistic wave equation is associated with a Lagrangian density, thus also with an energy-momentum tensorial density. The wave of an electron cloud adds these energy-momentum densities, while photons in light are precisely those differences between such energy-momentum densities.
