红外焦平面阵列的非均匀性研究  

Study of non-uniformity in infrared focal plane arrays

作　　者：王学伟[1] 李珂[1] 王世立[1] 

机构地区：[1]海军航空工程学院,山东烟台264001

出　　处：《激光与红外》2012年第10期1131-1134,共4页Laser & Infrared

基　　金：国家教育部新世纪优秀人才支持计划项目资助

摘　　要：红外焦平面阵列的非均匀性校正是凝视型红外成像系统中必不可少的一部分。文中从光电成像原理角度出发,深层分析了非均匀性的产生机理;针对传统神经网络校正算法存在的缺陷,从理想期望输出、误差函数和迭代步长三个方面进行了改进;并通过实验仿真对改进算法进行了验证,结果表明改进算法的校正效果较传统算法有明显的提高。The non-uniformity correction of infrared focal plane arrays is the essential part in the stare type infrared system. The article analyses the formation mechanism of non-uniformity from the angle of optical imaging principle. According to the limitation of traditional neural network algorithm, it makes improvement from ideal expectation out put, error function and iterate step. The improved algorithm is verified by the simulation experiment and the result shows that the new algorithm has obvious advantage in correction effect.

关 键 词：红外焦平面阵列 非均匀性 神经网络算法 

分 类 号：TN21[电子电信—物理电子学] TP391[自动化与计算机技术—计算机应用技术]