| 基于人工神经网络的红外焦平面阵列定标方法 |
| |
| 引用本文： | 张云涛,李涛,高太长. 基于人工神经网络的红外焦平面阵列定标方法[J]. 气象水文海洋仪器, 2009, 26(4): 19-21,25 |
| |
| 作者姓名： | 张云涛 李涛 高太长 |
| |
| 作者单位： | 解放军理工大学气象学院,南京,211101 |
| |
| 摘 要： | 用标准黑体定标数据分别采用基于神经网络和基于线性模型的定标方法对非制冷红外焦平面阵列上单个单元进行了定标，并对二者定标结果进行了分析和讨论。基于线性模型的定标方法采用最小二乘法拟合定标系数，数据处理简单，但精度较低。人工神经网络定标方法是在未知的情况下通过设计合适的网络就可以逼近输入与输出之间的映射关系，提高了定标精度。
|
| 关 键 词： | 非制冷红外焦平面阵列 辐射定标 神经网络 最小二乘法 |
Calibration method of infrared focal plane arrays based on neural network |
| |
| Zhang Yuntao,Li Tao,Gao Taichang. Calibration method of infrared focal plane arrays based on neural network[J]. Meteorological，Hydrological and Marine Instruments, 2009, 26(4): 19-21,25 |
| |
| Authors: | Zhang Yuntao Li Tao Gao Taichang |
| |
| Affiliation: | ( Institute of Meteorology, PLA University of Science and Technology, Nanjing 211101 ) |
| |
| Abstract: | The single pixel of UIRFPA（uncooled infrared focal plane arrays） were calibrated by using the calibrating data of standard radiant source based on neural network method and linear model method. The calibrating results of the two methods were analyzed and discussed. Linear model method give the calibration coefficient by least square algorithm,which is simple and low accuracy. The neural network method can simulate the function of input and output value by designing a right network structure at unknown situation,which can improve the calibration results. |
| |
| Keywords: | UIRFPA radiant calibration neural network least square algorithm |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
