摘要
长波红外高光谱系统既可分析目标成分又能感应其温度信息,广泛应用于遥感领域。该系统响应波段为7.7~9.5μm,共32个波段,光谱分辨率达53 nm,可用于地质勘探,环境监测等领域。由于长波红外器件工艺问题,探测器在成像之前必须进行非均匀性校正。通过对比基于焦平面和基于光谱的校正方法,说明后者能够很好地保持目标的原始光谱特征,并且不具有场景依赖的特点。在此基础上,提出从原始光谱曲线中提取目标光谱特征的方法,得到不同温度黑体和高光谱图像上不同目标的光谱曲线。
LW infrared hyperspectral system can analyze target composition as well as sense its temperature,therefore it is widely used in remote sensing environment.The presented system collects hyperspectral imagery from 7.7 to 9.5 μm in 32 spectral bands and can be used in the detection of geological exploration,environmental monitoring,etc.The pixel responses of LW infrared FPA（Focal Plane Array） are uneven because of immature FPA techniques,so the ununifomity correction must be conducted in original data.LW infrared hyperspectral image includes the scene of the target as well as the spectral feature of the target.Through the comparison of ununiformity correction algorithms based on IRFPA and spectral feature,it is shown that the latter algorithm can keep target′s original spectral feature and be non-scene-dependent.Furthermore,the method is proposed to extract target′s spectral feature through original spectral curve and get spectral curve of blackbody in different temperatures and different targets on the hyperspectral image.
出处
《红外与激光工程》
EI
CSCD
北大核心
2011年第4期605-610,共6页
Infrared and Laser Engineering
基金
国家863计划(2007AA12Z104)
国家973计划(2009CB724005)
关键词
长波红外高光谱
非均匀性校正
光谱特征提取
LW infrared hyperspectral
nonuniformity correction
spectral feature abstraction
