Image-driven evaluation metric for a space-based infrared diurnal detection analysis for flying aircrafts

Author:

Jia Guorui,Li Jiyuan,Luo Weizhong,Zhao Huijie1

Affiliation:

1. Beihang University

Abstract

The traditional evaluation calculation method fails to adequately consider the impact of image degradation and ignores the uncertainty caused by instrument noise on point target detectable status. This overestimates the detection capability of the system and cannot meet the needs of the point target detection evaluation under space-based observations. An image-driven evaluation metric (IDEM) is proposed in this paper, considering the effects of image degradation on the target, the background, and its clutter signals while providing coefficients of variation (CVs) for evaluation metrics. Image sequences of aircraft in different bands of the infrared imaging system were generated for cross validation. A comparative analysis against the traditional method shows that instrument noise significantly interferes with the point target signal in the image, and our method provides a more accurate and comprehensive evaluation of aircraft detectability under space-based infrared observations. Based on this, the IDEM maps under day and night in the mid- and long-infrared were computed and analyzed under space-based infrared observations of the flying aircraft. The results indicate that the mid-wave infrared (MWIR) is more sensitive to diurnal variation, whereas the long-wave infrared (LWIR) is more stable but has a lower daytime detectability compared to mid-infrared. Theoretically, both mid- and long-infrared enable night detection capabilities for point targets under the proper conditions, with mid-infrared offering higher detection potential. Our works provide new insight and approaches for the point target detection evaluation, system design, and on-orbit testing.

Funder

National Key Research and Development Program of China

Publisher

Optica Publishing Group

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3