摘要
借鉴生物免疫机制,提出一种基于免疫算法的自动目标识别性能与热成像分辨率关系的分析方法,研究热红外目标在不同分辨等级(发现、识别、确认)所需的热成像空间分辨率。该方法以红外多分辨率目标作为研究对象,根据免疫机制的反向选择原理计算各分辨率对应的配准概率,并与设定门限比较,从而给出热红外目标在各任务等级的最低分辨率要求。最后以计算机仿真的红外目标为例进行了验证。
In order to research desired imaging resolution of thermal target under different discrimination levels, including detection, recognition, and identification, by referring to biological immune mechanism, a new method based on immune algorithm, which is used to analyze the relation between target recognition performance and thermal imaging resolution, is proposed. Taking infrared multi-resolution targets as the study objects, we compute registration probability under different resolution levels based on negative selective theory of im- mune mechanism, and compare with the set thresholds, so the desired lowest imaging resolutions of thermal target under different task grades are obtained. Finally, taking a computer simulated infrared target for example, the experiment results are given.
出处
《计算机与数字工程》
2013年第12期1970-1972,共3页
Computer & Digital Engineering
基金
湖北省自然科学基金资助项目(2012FFB04711)资助
关键词
热成像
空间分辨率
目标识别性能
免疫算法
thermal imaging, spatial resolution, performance of target recognition, immune algorithm
