摘要
基于光学图像分析的目标伪装效果评估对于提高现役主战装备在战场上的生存能力有重要的指导意义,是军事国防领域中研究的热点问题之一。为了给目标探测与识别、伪装方案设计等相关领域研究学者提供参考,对传统光学伪装效果评估模型、神经网络评估模型、基于人眼视觉注意机制模型以及多属性决策模型进行了介绍,阐述了目前国内外典型评估方法的基本思路及应用,分析了不同模型的特点和不足以及光学伪装效果评估面临的问题。在此基础上,指出模型训练样本的获取和模型参数的自适应性以及运动目标伪装评估的时效性是未来伪装效果评估的研究重点。
Evaluation of camouflage effectiveness based on optical image analysis has an important guiding significance to improve the survivability of main battle equipment in the battlefield, and it is one of the hot issues in the field of military defense. To facilitate researchers in related fields, the evaluation methods and research progress of optical camouflage effect are introduced briefly, such as conventional method, neural network evaluation model, human visual attention mechanism model and multi-attribute decision model. Then the basic ideas and applications of typical evaluation methods are expounded. After that, the characteristics and deficiencies of different models and the problems faced by the evaluation of optical camouflage effects are analyzed. On this basis, it is pointed out that the acquisition of training samples, the adaptability of model parameters and the timeliness of motion target camouflage evaluation are the research focuses of future camouflage effect evaluation.
作者
甘源滢
刘春桐
李洪才
马世欣
GAN Yuanying;LIU Chuntong;LI Hongcai;MA Shixin(Missile Engineering College,Rocket Force University of Engineering,Xi’an 710025,China)
出处
《应用光学》
CAS
CSCD
北大核心
2019年第6期1050-1058,共9页
Journal of Applied Optics
基金
军队“十三五”装备预研课题(30406020403)
关键词
兵器科学与技术
光学图像分析
伪装效果评估
armament science and technology
optical image analysis
camouflage effectiveness evaluation
