生成理解统一模型在军事指挥控制领域的应用综述

Review of the Applications of Generative Understanding Unified Models in Military Command and Control

李敏;燕松;张雨森;李睿璇;苟瑶;何玉杰; LI Min;YAN Song;ZHANG Yusen;LI Ruixuan;GOU Yao;HE Yujie;Rocket Force University of Engineering;

随着人工智能技术从单模态、判别式范式向多模态、生成式范式跨越，生成理解统一模型（generative understanding unified model, GUUM）作为一种新兴的技术架构，正在重塑现代军事力量的底层逻辑。首先，系统性地分析了GUUM的构建机理，揭示其通用能力的根源来自于多模态数据的统一建模与分析；然后，分析了其在感知层面应用的区别；最后，将视角放在物理战场的交互上，解析基于生成式规划与扩散策略的具身智能如何改变物理战场。在应用层面，结合具体案例深入探讨了GUUM在自动化作战方案生成与兵棋推演中的效能、利用GUUM进行情报融合的演进、合成孔径雷达图像的扩散模型增强技术，以及GUUM在认知战和电子战频谱生成中的战术价值。同时，基于最新对抗性攻击研究，揭示了统一模型在面对物理扰动、数据投毒及模型幻觉时的脆弱性。本文旨在为军事决策者及国防科研人员提供一份详尽的评估，论证GUUM如何成为推进新型作战概念由想定变为现实与全域指挥控制的关键赋能技术，并针对其安全风险提出防御性建议。

As artificial intelligence technology transitions from single-modal discriminative paradigms to multi-modal generative paradigms, the generative understanding unified model(GUUM)has emerged as a novel technical architecture that reshapes the underlying logic of modern military power. First, this study systematically analyzed the construction mechanisms of GUUM, revealing that its general capabilities stemmed from the unified modeling and analysis of multi-modal data. It further examined the distinctions in its application at the perception level. Finally, the review focused on interactions within the physical domain, analyzing how generative planning and diffusion policy-based embodied intelligence transformed the physical battlefield. At the application level, the review incorporated specific case studies to provide an indepth exploration of GUUM efficacy in the automated course of action generation and wargames, the evolution of intelligence fusion utilizing GUUM, diffusion model enhancement techniques for synthetic aperture radar images, and the tactical value of GUUM in cognitive and electronic warfare spectrum generation.Drawing on the latest research on adversarial attacks, this review exposed the vulnerabilities of unified models when facing physical perturbations, data poisoning, and model hallucinations. This study aimed to provide a comprehensive assessment for military decision-makers and defense researchers, demonstrating how GUUM served as a key enabling technology for advancing novel operational concepts from scenarios to reality and facilitating all-domain command and control while offering defensive recommendations to address security risks.

磁流变液的核心性能与应用研究综述

Review on the Core Properties and Applications of Magnetorheological Fluids

桑媛园;冯永保;张艺萱;韩小霞;邓宜琳; SANG Yuanyuan;FENG Yongbao;ZHANG Yixuan;HAN Xiaoxia;DENG Yilin;

磁流变液作为一种典型的智能可控材料，具有响应速度快、力学性能优异、控制方式简便等优势，在工程领域具有广泛的应用前景。本文系统梳理了近年来国内外有关磁流变液的研究成果。首先，对磁流变液的组成成分进行深入剖析；其次，着重阐述了磁流变液的稳定性、磁特性以及流变性能，并深入分析了其中需重点解决的沉降分层、温度适应性和剪切稀化等问题；最后，总结了磁流变液在减振缓冲、精密加工等工程中的典型应用，进一步提出了材料制备工艺优化和器件设计升级等未来发展方向。本文可为磁流变液的配方优化、性能调控与工程化应用提供参考。

As typical intelligent controllable materials, magnetorheological fluids possess advantages such as fast response speed, excellent mechanical properties, and simple control methods: thus they have broad application prospects in the engineering field. This paper systematically summarized the research achievements in magnetorheological fluids in recent years. First, an in-depth analysis of the composition of the magnetorheological fluids was conducted. Subsequently, this study focused on expounding the stability,magnetic properties, and rheological properties of magnetorheological fluids and conducted an in-depth analysis of the key issues to be addressed, including sedimentation and stratification, temperature adaptability, and shear thinning. Finally, typical applications of magnetorheological fluids in engineering fields, such as vibration reduction and precision machining, were summarized. Future development directions, such as optimization of material preparation processes and upgrading of device design, were proposed, with the aim of providing theoretical references and technical guidance for formula optimization, performance regulation,and engineering applications of magnetorheological fluids.

基于频谱弥散调制的间歇采样转发干扰方法

Interrupted Sampling Repeater Jamming Method Based on Smeared Spectrum Modulation

吴其华;张楷煜;徐志明;刘晓斌;顾赵宇;赵锋; WU Qihua;ZHANG Kaiyu;XU Zhiming;LIU Xiaobin;GU Zhaoyu;ZHAO Feng;

针对传统间歇采样转发干扰（interrupted sampling repeater jamming, ISRJ）方法存在假目标分布规律性强、高阶假目标峰值衰减速度快，导致ISRJ信号时频特征明显、易被雷达系统识别与抑制的问题，提出一种基于频谱弥散的间歇采样转发干扰（smeared spectrum-interrupted sampling repeater jamming, SMSP-ISRJ）方法。该方法将频谱弥散技术引入间歇采样转发机制中，通过对所截取的雷达局部信号片段进行相位重映射来改变其调频斜率，从而改变了干扰信号的脉压输出特性与时频分布规律。仿真结果表明：相较于传统ISRJ方法，SMSP-ISRJ产生的0阶假目标到1阶假目标的幅度衰减速率明显减缓，且通过调节采样周期与压缩因子等参数，可灵活实现欺骗干扰和压制干扰的效果切换；在抗识别与抗抑制方面，SMSP-ISRJ使基于时频域行聚类的抗干扰算法失效，干扰信号经抑制后的脉压输出在真实目标附近仍残留多个高幅值虚假峰。

To address the issues of traditional interrupted sampling repeater jamming(ISRJ), such as regular false target distribution and rapid peak attenuation of high-order false targets, which resulted in distinct time-frequency characteristics of ISRJ signals and vulnerability to identification and suppression by radar systems, a smeared spectrum-interrupted sampling repeater jamming(SMSP-ISRJ)method was proposed. This method introduced the smeared spectrum technique into the interrupted sampling repeater mechanism and remapped the phase of the intercepted local radar signal segments to alter the frequency modulation slope of the radar signal fragments, thereby changing the pulse compression output characteristics and time-frequency distribution patterns of jamming signals. Simulation results show that, compared with traditional ISRJ methods, the SMSP-ISRJ method significantly slows down the amplitude attenuation rate from the 0th- to 1st-order false targets. It can flexibly switch between deceptive and suppressive jamming effects by adjusting parameters, such as sampling periods and compression factors. In terms of antiidentification and anti-suppression, SMSP-ISRJ disables anti-jamming algorithms based on time-frequency row clustering. After suppression, multiple high-amplitude false peaks remain near the real target in the pulse compression output of the jamming signals.

基于MSNR-JBSS的雷达组网系统抗主瓣压制干扰方法

An Anti-Mainlobe Suppressive Jamming Method for Radar Networked Systems Based on MSNR-JBSS

韩晓斐;刘骋域;张琪;何华锋;周涛; HAN Xiaofei;LIU Chengyu;ZHANG Qi;HE Huafeng;ZHOU Tao;

针对现有抗主瓣压制干扰方法依赖干扰类型、干扰角度等先验信息的问题，提出一种基于最大信噪比的联合盲源分离（maximum signal-to-noise ratio for joint blind source separation, MSNR-JBSS）的雷达组网系统抗主瓣压制干扰方法。首先，该方法以组网系统各雷达站为基准雷达，对雷达接收信号进行延时估计与补偿；然后，基于MSNR准则进行联合盲源分离处理，获得以每部雷达为基准的目标径向距离信息；最终，基于多站联合定位的数据级融合方法实现目标精确定位。该方法虚拟扩展了信号通道数，增强了源信号的分离效果，且无需提前获取干扰类型、干扰角度、信号源个数等先验信息。仿真结果表明：所提方法抗干扰与定位性能均优于对比方法。在相同仿真条件下，所提方法的输出峰值旁瓣比平均低于对比方法 1.5～3 dB；在干噪比100 dB、信噪比10 dB的强干扰场景中，目标定位精度可达3.971 m，较对比方法提升显著，充分验证了通过JBSS虚拟扩展信号通道数提高抗干扰性能的有效性。

To address the reliance of existing anti-mainlobe suppressive jamming methods on prior information, such as jamming type and direction, a method for radar networks to counter mainlobe suppressive jamming based on the maximum signal-to-noise ratio for joint blind source separation(MSNR-JBSS)was proposed. First, each radar in the networked system was sequentially designated as the reference radar, and all the received radar signals were estimated and compensated. Subsequently, the JBSS processing was conducted based on the MSNR criterion to obtain the radial-range information of the target, with each radar serving as a reference. Finally, precise target localization was achieved using a data-level fusion approach based on multi-station joint localization. The proposed method virtually expanded the number of signal channels and enhanced the separation performance of source signals, requiring no prior information such as jamming type, jamming direction, or number of signal sources. Simulation results demonstrate that the proposed method outperforms the comparative methods in terms of both anti-jamming performance and localization accuracy. Under the same simulation conditions, the output peak-to-side lobe ratio is, on average, 1.5-3 dB lower than that of the benchmark methods. In a strong interference scenario with a jamming-to-noise ratio of 100 dB and a signal-to-noise ratio of 10 dB, the target localization accuracy reaches 3.971 m, showing a significant improvement over the compared approaches. These results verify the improvement in the anti-jamming performance achieved by virtually expanding the number of signal channels through the JBSS.

稀疏空间特征与线性注意力融合的对空红外小目标检测

Sparse Spatial Feature and Linear Attention Fusion-Based Aerial Infrared Small Target Detection

陈光晨;张枫;张印辉;何自芬;杨小冈;卢瑞涛; CHEN Guangchen;ZHANG Feng;ZHANG Yinhui;HE Zifen;YANG Xiaogang;LU Ruitao;

针对空中红外小目标探测存在目标距离远、热辐射衰减严重及环境信息复杂等问题，提出稀疏空间特征与线性注意力融合的对空红外小目标检测（aerial infrared small target detection,AISTD）模型。首先，为解决因特征图尺寸降低而导致目标细节信息丢失的问题，设计了稀疏空间特征提取模块和对称式挤压激励特征提取模块来加强模型对红外小尺度目标的特征提取能力；然后，针对热辐射信号易受环境干扰的问题，将二维图像转换为一维序列后输入到多头线性注意力模块，并对其进行旋转位置编码，从多个注意力子空间中获取红外目标细节纹理和轮廓特征信息，以增强模型在复杂场景下对红外小目标的关注度；最后，建立对空红外小目标数据集并对模型进行训练和测试，以验证模型的有效性。结果表明：所提方法在对空红外小目标数据集上的mAP75达到90.3%,mAP50-95达到74.7%，模型推理时间缩短至3.9 ms，相比现有的YOLOv11和YOLOv13等目标检测网络，AISTD能够更快速准确地识别出空中红外小目标。

Aerial infrared small-target detection faces critical challenges including long detection distances, severe thermal radiation attenuation, and complex environmental interference. To address these issues, an aerial infrared small-target detection(AISTD)model that effectively integrates sparse spatial features with linear attention was proposed. First, to alleviate the loss of fine-grained feature information caused by the reduced feature map, a sparse spatial feature extraction module and a symmetrical squeezing excitation feature extraction module were introduced to enhance the model's capability of infrared smallscale target feature extraction. Subsequently, to mitigate environmental interference with thermal radiation signals, the two-dimensional image was converted into a one-dimensional sequence and input into a multihead linear attention module, with rotary positional encoding applied. The proposed model captured detailed texture and contour information of infrared targets across multiple attention subspaces, thereby improving the sensitivity to small targets in complex aerial scenes. Furthermore, an aerial infrared small-target dataset was constructed to train and test the proposed approach and validate its effectiveness. Extensive experimental results demonstrate that the proposed AISTD achieves an mAP75 of 90.3% and an mAP50-95 of 74.7% on a self-built dataset, while reducing the inference time to 3.9 ms. Compared with existing object detection networks, such as YOLOv11 and YOLOv13, the AISTD model exhibits superior performance in terms of both detection accuracy and computational efficiency.

无人机航迹规划智能算法研究综述

Review of Intelligent Algorithms for Unmanned Aerial Vehicle Trajectory Planning

鲜勇;郭靖;任乐亮;李少朋; XIAN Yong;GUO Jing;REN Leliang;LI Shaopeng;

针对传统无人机航迹规划算法计算效率低、适应性差的问题，从算法改进的角度出发，系统地梳理了国内外在无人机航迹规划领域的研究进展。从全局优化能力、动态适应性、计算效率以及鲁棒性等方面，阐述了无人机航迹规划智能算法具有的显著优势；同时，对航迹规划智能算法进行了详细分类和总结，从智能优化算法与机器学习算法的几种典型算法改进策略出发，对比分析了每种典型航迹规划智能算法的优缺点、应用场景以及改进方向；最后，结合当前无人机航迹规划所面临的多重挑战，对无人机航迹规划智能算法未来的研究趋势和潜在应用方向进行了展望，强调了航迹规划智能算法在提升算法性能、适应复杂环境、满足多样化任务需求以及智能算法融合等方面的重要性，为无人机航迹规划智能算法的研究与发展提供了参考。

To address the problems of low computational efficiency and poor adaptability of conventional unmanned aerial vehicle(UAV)trajectory planning algorithms, progress in domestic and international research in this field was reviewed from the perspective of algorithm improvement. First, the substantial advantages of intelligent algorithms for UAV trajectory planning were elaborated based on global optimization capability, dynamic adaptability, computational efficiency,and robustness. Subsequently, a detailed classification and summary of intelligent algorithms for trajectory planning were presented. Furthermore, the advantages, disadvantages, application scenarios, and improvement directions of each typical intelligent algorithm were comparatively analyzed according to typical algorithm improvement strategies of intelligent optimization and machine learning algorithms. Finally, combined with the multiple challenges currently encountered in UAV trajectory planning, future research trends and potential application directions of intelligent algorithms for UAV trajectory planning were discussed. Additionally, the importance of intelligent algorithms for trajectory planning in improving algorithm performance, adapting to complex environment, meeting diverse task requirements, and integrating intelligent algorithms were emphasized. The review provides a reference for the research and development of intelligent algorithms for UAV trajectory planning.

高温后钢纤维混凝土的力学性能与微观结构

Mechanical Properties and Microstructure of Elevated Temperature-processed Steel Fiber Reinforced Concrete

董婧;杨进良;王炜;刘渊;杜娟; DONG Jing;YANG Jinliang;WANG Wei;LIU Yuan;DU Juan;

为研究不同温度作用对钢纤维混凝土力学性能与微观结构的影响，对比了经25℃、200℃、400℃、600℃温度处理后自然冷却的普通混凝土与体积掺量为1%的钢纤维混凝土的力学性能与微观结构。首先，通过混凝土立方体抗压试验及劈裂抗拉试验获取了试件的破坏形态、抗压强度、劈裂抗拉强度；然后，使用霍普金森压杆装置开展了4种冲击气压下试件的动态冲击压缩试验，获取了动态抗压强度与动态压缩增长因子，并运用扫描电镜观察了不同处理温度条件下试件的微观结构。研究结果表明：随着处理温度的升高，普通混凝土与钢纤维混凝土的颜色、表观裂纹、质量以及微观结构上均会产生变化，抗压强度与劈裂抗拉强度均呈降低趋势，且在400～600℃时出现了骤降；钢纤维的加入可以在一定程度上降低混凝土的高温后质量损失率，有效改善混凝土的动静态力学性能。此外，钢纤维的桥接作用可在一定程度上抑制混凝土裂纹的扩展。

The effects of different temperatures on the mechanical properties and microstructure of steel fiber reinforced concrete(SFRC)was investigated by conducting a comparative study of the mechanical properties and microstructure of normal concrete(NC)that wasnaturally cooled after elevated-temperature treatment at 25 ℃, 200 ℃, 400 ℃ and 600 ℃, and SFRC with a volume proportion of 1%. First, the failure mode, compressive strength, and splitting tensile strength were obtained by compression and splitting tensile tests of a concrete cube. Second, the dynamic compressive strength and dynamic compression growth factors were obtained through dynamic impact compression tests, which were conducted on the specimens under four different impact pressures using Split Hopkinson Pressure Bar. The micro-structures of the specimens under different temperature treatments were observed using scanning electron microscope. Results showed that with an increase in temperature, both NC and SFRC change in terms of color, surface cracks, mass, and microstructure. Simultaneously, the compression and splitting tensile strengths show a decreasing trend, with a sudden drop at 400～600 ℃. However, the addition of steel fibers can reduce the mass-loss rate of concrete after high-temperature treatment, and effectively improve the dynamic and static mechanical properties of concrete. Additionally, the bridging effect of the steel fibers can inhibit the propagation of concrete cracks to some extent.

基于目标HSV（色调-饱和度-亮度）空间图像自适应分割的直线检测算法

Straight Line Detection Algorithm Based on Adaptive Segmentation of Target HSV Space Images

尉金强;杜文正;孙晓艳;聂婷;唐圣金; WEI Jinqiang;DU Wenzheng;SUN Xiaoyan;NIE Ting;TANG Shengjin;

针对现阶段利用数字图像进行半挂车倒车辅助过程中，存在辅助地标线检测效果鲁棒性差和实时性不高等问题，提出了一种基于目标色调-饱和度-亮度（Hue-SaturationValue,HSV）空间图像自适应分割的直线检测算法。首先，对不同场景下的直线目标进行分割提取，通过统计不同场景下目标HSV空间图像直方图确定图像分割的初始阈值，并基于初始阈值提出一种基于目标HSV空间的图像自适应分割算法。然后，对图像进行预处理及边缘检测，提取目标边缘点。最后，采用基于中位数绝对偏差的最小二乘直线检测方法剔除边缘点中的离群值以提高地标直线目标的检测精度。实验结果表明：该算法能够在不同背景、不同光照条件下对目标进行准确检测，具有良好的鲁棒性；与Hough变换算法相比，该算法的运算速度更快，直线检测的准确率更高，并且对不同颜色地标线的检测也具有良好的适用性。

To solve problems in exisiting digital images processing-based semi-trailer reversing assistant system development, such as poor robustness and low real-time of landmark line detection, a straight line detection algorithm based on adaptive segmentation of target Hue-Saturation-Value(HSV) space images was proposed. First, line targets in various scenes were segmented and extracted. Simultaneously, histograms of target images in the HSV space across different scenes were used to determine the initial threshold for image segmentation, based on which an image adaptive segmentation algorithm according to target HSV space images was proposed.Subsequently, the preprocessing and edge detection of images were conducted to extract edge points of targets. Finally, outliers of edge points were eliminated by using a median absolute deviation-based least squares straight line detection method to improve the detection accuracy of landmark line targets. Experimental results showed that the proposed algorithm is able to accurately detect targets under different backgrounds and lighting conditions with good robustness.Compared to the Hough transform algorithm, the algorithm has faster operation speed, higher accuracy of straight line detection, and good applicability to the detection of landmark lines of different colors.

偏振探测成像技术研究进展及其军事应用

Progress and Military Applications of Polarization Detection Imaging Technology

李林;张国清;周战荣;耿燚; LI Lin;ZHANG Guoqing;ZHOU Zhanrong;GENG Yi;

对代表性的偏振探测技术及其在军事领域中的主要应用给出了总结。首先,介绍了利用琼斯矩阵和穆勒矩阵进行偏振探测的基本原理；其次，就4种常见的偏振探测技术的实现方法及其优缺点进行了讨论，包括分时偏振成像技术和同时偏振成像技术（包括分振幅偏振成像、分孔径偏振成像、分焦平面偏振成像技术）；然后，介绍了偏振探测技术在目标探测、图像去雾、导航、地雷探测等军事方面的典型应用；最后,指出了偏振探测技术目前存在的难点问题以及未来发展展望。

Representative polarization detection technologies and its applications in military were summarized. Firstly, the basic principles of polarization detection using Jones matrix and Muller matrix were introduced. Secondly, four common polarization detection techniques, including the division-of-time polarization imaging as well as the division-of-amplitude polarization imaging, the aperture-divided polarization imaging and the division-of-focal-plane polarization imaging of the simultaneous polarization imaging, and their advantages and disadvantages were discussed. Then, typical applications of polarization detection technology in the military field, such as target detection, image dehazing, navigation and landmine detection, were introduced. Finally,the current difficulties and an outlook on the development of polarization detection technology were provided.

重型车辆高级驾驶辅助系统研究综述

Review on ADAS Technologies for Heavy-Duty Vehicles

张志利;祝慧鑫;周召发;赵军阳;张克凡; ZHANG Zhili;ZHU Huixin;ZHOU Zhaofa;ZHAO Junyang;ZHANG Kefan;

针对高级驾驶辅助系统（Advanced Driving Assistance System,ADAS）重型车辆视野盲区大、反应要求高、制动时间短的安全特征，论述了重型车辆ADAS技术的发展历程以及目前ADAS技术的涵盖范围，重点论述了重型车辆ADAS的环境感知技术（包括基于视觉类图像、雷达类点云、多模态融合的目标检测技术）以及定位定向技术（囊括局部高精地图构建和多模态融合定位定向方案研究）。展望了重型车辆ADAS技术的发展趋势。

Considering safety characteristics of heavy-duty vehicles such as large blind spots,high reaction requirements and short braking time, the development courses of heavy-duty vehicle ADAS technologies and the current coverage of ADAS technologies were reviewed. First, environment sensing technologies of heavy-duty vehicle ADAS, including target detection technologies based on vision-like images, radar-like point clouds and multimodal fusion were discussed respectively. Second, the localization and orientation technology of heavy-duty vehicle ADAS,encompassing the construction of local high-precision maps and the research of multimodal fusion localization and orientation scheme were described. Finally, the trend of further development of ADAS technologies for heavy-duty vehicles were predicted.

无人机航迹规划智能算法研究综述

Review of Intelligent Algorithms for Unmanned Aerial Vehicle Trajectory Planning

鲜勇;郭靖;任乐亮;李少朋; XIAN Yong;GUO Jing;REN Leliang;LI Shaopeng;

针对传统无人机航迹规划算法计算效率低、适应性差的问题，从算法改进的角度出发，系统地梳理了国内外在无人机航迹规划领域的研究进展。从全局优化能力、动态适应性、计算效率以及鲁棒性等方面，阐述了无人机航迹规划智能算法具有的显著优势；同时，对航迹规划智能算法进行了详细分类和总结，从智能优化算法与机器学习算法的几种典型算法改进策略出发，对比分析了每种典型航迹规划智能算法的优缺点、应用场景以及改进方向；最后，结合当前无人机航迹规划所面临的多重挑战，对无人机航迹规划智能算法未来的研究趋势和潜在应用方向进行了展望，强调了航迹规划智能算法在提升算法性能、适应复杂环境、满足多样化任务需求以及智能算法融合等方面的重要性，为无人机航迹规划智能算法的研究与发展提供了参考。

To address the problems of low computational efficiency and poor adaptability of conventional unmanned aerial vehicle(UAV)trajectory planning algorithms, progress in domestic and international research in this field was reviewed from the perspective of algorithm improvement. First, the substantial advantages of intelligent algorithms for UAV trajectory planning were elaborated based on global optimization capability, dynamic adaptability, computational efficiency,and robustness. Subsequently, a detailed classification and summary of intelligent algorithms for trajectory planning were presented. Furthermore, the advantages, disadvantages, application scenarios, and improvement directions of each typical intelligent algorithm were comparatively analyzed according to typical algorithm improvement strategies of intelligent optimization and machine learning algorithms. Finally, combined with the multiple challenges currently encountered in UAV trajectory planning, future research trends and potential application directions of intelligent algorithms for UAV trajectory planning were discussed. Additionally, the importance of intelligent algorithms for trajectory planning in improving algorithm performance, adapting to complex environment, meeting diverse task requirements, and integrating intelligent algorithms were emphasized. The review provides a reference for the research and development of intelligent algorithms for UAV trajectory planning.

偏振探测成像技术研究进展及其军事应用

Progress and Military Applications of Polarization Detection Imaging Technology

李林;张国清;周战荣;耿燚; LI Lin;ZHANG Guoqing;ZHOU Zhanrong;GENG Yi;

对代表性的偏振探测技术及其在军事领域中的主要应用给出了总结。首先,介绍了利用琼斯矩阵和穆勒矩阵进行偏振探测的基本原理；其次，就4种常见的偏振探测技术的实现方法及其优缺点进行了讨论，包括分时偏振成像技术和同时偏振成像技术（包括分振幅偏振成像、分孔径偏振成像、分焦平面偏振成像技术）；然后，介绍了偏振探测技术在目标探测、图像去雾、导航、地雷探测等军事方面的典型应用；最后,指出了偏振探测技术目前存在的难点问题以及未来发展展望。

Representative polarization detection technologies and its applications in military were summarized. Firstly, the basic principles of polarization detection using Jones matrix and Muller matrix were introduced. Secondly, four common polarization detection techniques, including the division-of-time polarization imaging as well as the division-of-amplitude polarization imaging, the aperture-divided polarization imaging and the division-of-focal-plane polarization imaging of the simultaneous polarization imaging, and their advantages and disadvantages were discussed. Then, typical applications of polarization detection technology in the military field, such as target detection, image dehazing, navigation and landmine detection, were introduced. Finally,the current difficulties and an outlook on the development of polarization detection technology were provided.

基于三维高斯泼溅技术的航拍场景建模

3D Gaussian Splatting-based Aerial Video Scene Modeling

何天琪;宋佳洁;程景春;潘雄; HE Tianqi;SONG Jiajie;CHENG Jingchun;PAN Xiong;

针对航拍视频场景建模存在的数据匮乏、现有算法建模细节缺失等问题，提出了基于仿真引擎的飞行场景数据采集方法，并采用三维高斯泼溅（3D Gaussian Splatting,3DGS）算法实现航拍场景的建模与渲染。首先，利用仿真引擎实现城区、沙滩、江桥3种典型场景下不同飞行轨迹、不同飞行速度、不同环境条件下的数据集获取，并进行航拍视频帧采样；然后，通过运动恢复结构实现场景的稀疏建模，再利用3DGS算法进行场景的稠密建模及渲染；最后，进行了模型的定量对比与定性分析实验。结果表明：与经典的COLMAP算法相比，所提算法具有更好的建模效果以及更完备的图像细节；线性、网格、螺旋不同的飞行轨迹均能得到较高的图像质量，但螺旋轨迹具有更好的场景通用性；飞行速度在30 m/s以下时可以得到较高质量的建模效果，其中PSNR>32 dB,SSIM>0.93,LPIPS<0.13；在不同环境条件下均能得到较高的图像质量，且在晴天白天理想光照下质量最佳。该研究为具有位姿灵活、飞行速度较快、场景大等难点的航拍视频提供了高精度、高鲁棒性的建模方案。

To solve the problems of data scarcity and lack of modeling details in existing algorithms in the modeling of aerial video scenes, a flight scene data acquisition method based on simulation engines was proposed. The advanced 3DGS(3D Gaussian Splatting)algorithm was adopted to model and render aerial scenes. First, simulation engines were used to obtain datasets for different flight trajectories, speeds, and environmental conditions in three typical scenes(urban areas, beaches, and river bridges), and to obtain samples of aerial video frames. In addition, the sparse modeling of the scenes was realized through the motion recovery structure, and the 3DGS algorithm was used for dense modeling and rendering of the scenes. Finally, quantitative comparison and qualitative analysis experiments of the models were conducted. The results show that compared with the classical COLMAP algorithm, the proposed algorithm has better modeling results and more complete image details. Higher image quality can be obtained in different flight trajectories(linear, grid, and spiral), while achieving better scene universality in spiral trajectories. Higher-quality modeling results can be obtained when the flight speed is below30 m/s(PSNR>32 dB, SSIM>0.93, LPIPS<0.13). Furthermore, high image quality can be achieved under different environmental conditions, and the best quality is achieved under ideal light during the daytime of sunny days. A high-precision and high-robustness modeling solution for aerial videos with difficulties such as flexible positions, fast flight speeds, and large scenes is provided in this paper.

一种电荷灵敏前置放大器的设计与实现

Design and Implementation of a Charge Sensitive Preamplifier

王佳帅;秦晋;许鹏;卜洪涛;王道平; WANG Jiashuai;QIN Jin;XU Peng;BU Hongtao;WANG Daoping;Xi'an CNNC Nuclear Instrument Co.Ltd;

针对传统金硅面垒α谱仪所使用的电荷灵敏前置放大器存在电路结构复杂、损坏后维修周期长的问题，基于低噪声场效应管和集成运算放大器，使用电路仿真软件，设计了针对金硅面垒半导体探测器的电荷灵敏前置放大器，简化了电路结构。对该前置放大器参数进行了理论计算和实验测试。结果表明：自制的电荷灵敏前置放大器的电荷变换增益为5×10~12V/C，能量变换增益为222 mV/MeV，上升时间为200 ns；将其应用于金硅面垒α谱仪之中，能够测量~239Pu板状源的α能谱。

To cover shortages of the charge sensitive preamplifier used in the existing Au-Si surface barrier alpha spectrometer, such as complex circuit structure and long maintenance cycle after damage, a charge sensitive preamplifier for gold silicon surface barrier semiconductor detector was designed based on low noise field effect transistor and integrated operational amplifier.The proposed preamplifier was designed with circuit simulation software and its simpler circuit structure. Furthermore, parameters of the charge sensitive preamplifier was obtained by theoretical calculation and verified by experimental tests. Results showed that the charge conversion gain of the self-made charge sensitive preamplifier is 5×10~12V/C, energy conversion gain is 222 mV/MeV and rise time is 200 ns. When it was applied to the Au-Si surface barrier α spectrometer, it can measure the α energy spectrum of ~239Pu plated source.

基于目标HSV（色调-饱和度-亮度）空间图像自适应分割的直线检测算法

Straight Line Detection Algorithm Based on Adaptive Segmentation of Target HSV Space Images

尉金强;杜文正;孙晓艳;聂婷;唐圣金; WEI Jinqiang;DU Wenzheng;SUN Xiaoyan;NIE Ting;TANG Shengjin;

针对现阶段利用数字图像进行半挂车倒车辅助过程中，存在辅助地标线检测效果鲁棒性差和实时性不高等问题，提出了一种基于目标色调-饱和度-亮度（Hue-SaturationValue,HSV）空间图像自适应分割的直线检测算法。首先，对不同场景下的直线目标进行分割提取，通过统计不同场景下目标HSV空间图像直方图确定图像分割的初始阈值，并基于初始阈值提出一种基于目标HSV空间的图像自适应分割算法。然后，对图像进行预处理及边缘检测，提取目标边缘点。最后，采用基于中位数绝对偏差的最小二乘直线检测方法剔除边缘点中的离群值以提高地标直线目标的检测精度。实验结果表明：该算法能够在不同背景、不同光照条件下对目标进行准确检测，具有良好的鲁棒性；与Hough变换算法相比，该算法的运算速度更快，直线检测的准确率更高，并且对不同颜色地标线的检测也具有良好的适用性。

To solve problems in exisiting digital images processing-based semi-trailer reversing assistant system development, such as poor robustness and low real-time of landmark line detection, a straight line detection algorithm based on adaptive segmentation of target Hue-Saturation-Value(HSV) space images was proposed. First, line targets in various scenes were segmented and extracted. Simultaneously, histograms of target images in the HSV space across different scenes were used to determine the initial threshold for image segmentation, based on which an image adaptive segmentation algorithm according to target HSV space images was proposed.Subsequently, the preprocessing and edge detection of images were conducted to extract edge points of targets. Finally, outliers of edge points were eliminated by using a median absolute deviation-based least squares straight line detection method to improve the detection accuracy of landmark line targets. Experimental results showed that the proposed algorithm is able to accurately detect targets under different backgrounds and lighting conditions with good robustness.Compared to the Hough transform algorithm, the algorithm has faster operation speed, higher accuracy of straight line detection, and good applicability to the detection of landmark lines of different colors.