短波红外波段高温目标识别的可行性分析

Feasibility analysis of shortwave infrared band for recognition of high temperature target

目前主要采用热红外遥感数据识别高温目标，研究发现短波红外数据也具有较好的高温目标识别能力。在高温目标混合像元中，高温地物的温度和面积是待求解的未知数，也足决定混合像元波谱特性的重要参数。首先辏于能量守恒原理建立了地表高温目标混合像元的辐射能量方程，然后进行方程参数的敏感性分析。研究结果表明，高温目标面积百分比和常温地物反射率对高温地物的温度和面积反演最为敏感。以山西保德和陕西府符研究区ETM’第7波段的数据为例进行了高温目标识别研究，所识别的火点像元辐射通量密度为背景的1．36～4．76倍。经野外验汪，用马氏距离法识别高温目标的精度达到88％，表明短波红外波段的遥感数据可用于高温H标的以别。

At present, high temperature target recognition mainly uses thermal infrared remote sensing data. In this study, the authors have found that shortwave infrared band has better recognition capability than thermal infrared band. In the mixed pixel of high temperature targets, temperature and area of high temperature objects are unknown. They are the key parameters that can determine the spectral character of mixed pixels. Based on the constant energy principle, the authors formulated the radiation energy equation for the mixed pixel of high temperature targets on the Earth＇ s surface. The results of the sensibility analysis for the equation parameters show that the area percentage of high temperature targets and the reflection of the normal temperature targets are most sensitive to the invertion of the temperature and the area of high temperature targets. ETM ＋7 data obtained in Baode of Shanxi and Fugu of Shaanxi were used for study of high temperature target recognition. The radiation flux density of the recognized fire is about 1.36 to 4.76 times that of the background value. Field verification shows that Mahalanobis method has the precision of 88% , suggesting that shortwave infrared band can be used to recognize high temperature targets.