红外遥感模型的辐射场景图高分辨率研究  

High-resolution study of radiation scene map of infrared remote sensing model

作　　者：李志梅[1] 肖德贵[2] LI Zhimei;XIAO Degui(Guilin University of Aerospace Technology,Guilin Guangxi 541004,China;Hunan University,Changsha 410022,China)

机构地区：[1]桂林航天工业学院,广西桂林541004 [2]湖南大学,长沙410022

出　　处：《激光杂志》2020年第8期119-123,共5页Laser Journal

基　　金：桂林市人才平台建设项目(No.20190206-2);广西高校中青年教师基础能力提升项目(No.2018KY0651)。

摘　　要：针对红外遥感模型的辐射场景图应用过程中需要变焦的需求,研究红外遥感模型的辐射场景图高分辨率,采用TITAN模型建立红外遥感模型,分析红外遥感模型的辐射场景传输过程,并通过Cool-Torrance模型赋予红外遥感模型表面细节信息,选取基于熵变分的图像分辨率增强技术增强红外遥感模型辐射场景图分辨率,将原始红外遥感模型辐射场景图通过离散熵变分模型滤波,降低图像噪声;将新的行或列插入去噪后图像相邻行或列中,重复上述步骤实现辐射场景图放大,最终放大后图像即高分辨率辐射场景图。实验结果表明,采用该方法获取的红外遥感模型高分辨率辐射场景图峰值信噪比均高于33 dB,且具有较优的视觉质量。In response to the zooming needs in application of radiation scene map in infrared remote sensing model,the high resolution of radiation scene map of infrared remote sensing model is studied.The infrared remote sensing model is established using TITAN model,and radiation scene transmission process of infrared remote sensing model is analyzed.The Cool-Torrance model gives surface detail information to infrared remote sensing model.The image resolution enhancement technology based on entropy variation is selected to enhance the resolution of infrared remote sensing model radiation scene map.The original infrared remote sensing model radiation scene map is filtered by discrete entropy variation model to reduce image noise.A new row or column into adjacent row or column of denoised image is inserted,the above steps to enlarge radiation scene image are repeated,and the image,a high-resolution radiation scene image,is enlarged finally.The experimental results show that the peak signal-to-noise ratio of high-resolution radiation scene image of infrared remote sensing model obtained by this method is higher than 33dB,and has better visual quality.

关 键 词：红外 遥感 模型 辐射 场景图 分辨率 增强 熵变分 

分 类 号：TN299[电子电信—物理电子学]