首页 | 本学科首页   官方微博 | 高级检索  
     

基于特征综合的红外动态伪装效果评估
引用本文:应家驹,何永强,陈玉丹,周冰,武东生.基于特征综合的红外动态伪装效果评估[J].半导体光电,2019,40(3):407-411.
作者姓名:应家驹  何永强  陈玉丹  周冰  武东生
作者单位:陆军工程大学石家庄校区光学工程教研室,石家庄,050003;陆军工程大学石家庄校区光学工程教研室,石家庄050003;吉林大学珠海学院电子信息工程学院,广东珠海519041
摘    要:针对红外动态伪装系统,提出了一种基于特征综合的红外动态伪装效果评估方法。根据红外侦测图像的特性,选择了亮度对比度、直方图特征、纹理特征和边缘形状特征等四种特征指标,分别统计背景图像与伪装图像之间四种特征的相似度。四种特征相似度在综合评估过程中所占的比重不固定,特征权重分配依据伪装图像特征序列数据的信息熵的大小决定。最后累加单幅伪装图像的四种特征值与特征权重的乘积,作为该幅图像的综合评价结果。在红外动态伪装系统上进行了试验,通过采集背景图像和序列伪装图像,验证了效果评估算法的可行性和有效性。红外动态伪装效果的评估结果可为优化伪装设计算法提供数据支撑。

关 键 词:红外伪装  特征综合  权重分配  效果评估
收稿时间:2018/11/27 0:00:00

Evaluation of Infrared Dynamic Camouflage Effect Based on Feature Synthesis
YING Jiaju,HE Yongqiang,CHEN Yudan,ZHOU Bing and WU Dongsheng.Evaluation of Infrared Dynamic Camouflage Effect Based on Feature Synthesis[J].Semiconductor Optoelectronics,2019,40(3):407-411.
Authors:YING Jiaju  HE Yongqiang  CHEN Yudan  ZHOU Bing and WU Dongsheng
Affiliation:Dept.of Optical Engin., Shijiazhuang Campus of Army Engin.University, Shijiazhuang 050003, CHN,Dept.of Optical Engin., Shijiazhuang Campus of Army Engin.University, Shijiazhuang 050003, CHN;Dept.of Electronic Information Engin., Jilin University Zhuhai College, Zhuhai 519041, CHN,Dept.of Optical Engin., Shijiazhuang Campus of Army Engin.University, Shijiazhuang 050003, CHN,Dept.of Optical Engin., Shijiazhuang Campus of Army Engin.University, Shijiazhuang 050003, CHN and Dept.of Optical Engin., Shijiazhuang Campus of Army Engin.University, Shijiazhuang 050003, CHN
Abstract:The evaluation of camouflage effect is an important part of camouflage design. Aiming at the infrared dynamic camouflage system, an evaluation method of infrared dynamic camouflage effect based on feature synthesis is proposed. According to the characteristics of infrared detection images, four feature indexes, namely brightness contrast, histogram, texture and edge shape, were selected to calculate the similarity between background image and camouflage image. The similarity proportion of the four features in the comprehensive evaluation process is not fixed, and the feature weight distribution is based on the entropy of the disguised image feature sequence data. Finally, the product of four characteristic values and characteristic weights of a single camouflage image is added as the comprehensive evaluation result of the image. Experiments were carried out on the infrared dynamic camouflage system, and the feasibility and effectiveness of the algorithm were verified by collecting background images and sequence camouflage images. The results of dynamic infrared camouflage effect evaluation can provide data support for the optimization of camouflage design algorithm.
Keywords:infrared camouflage  feature synthesis  weight distribution  effect evaluation
本文献已被 万方数据 等数据库收录!
点击此处可从《半导体光电》浏览原始摘要信息
点击此处可从《半导体光电》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号