机载短波红外分光谱扫描数据进行油气藏勘查的模式识别研究

本文通过图像辐射恢复处理消除了大范围航空短波红外分光谱扫描图像的辐射畸变。以土壤吸附烃的光谱特性为依据，通过对训练区图像的因子分析，在主因子空间上，确定出土壤吸附烃异常的特征方向。通过确定方向因子分析的处理，实现了大范围的油气异常信息的识别、提取与分类。

Pattern Recognition of Oil Pool with SWIR Scanner Data

Hydrocarbon content anomalies in soil above oil pool are good directindicating index on the earth surface. Oil remote sensing anomalies were obtainedby airborne SWIR scanner images through micro-computer image processing,inclu-ding image restoration,oil anomalous information extraction and recognition basedon spectral features of the index. The result was proved to be correspondent with the ascertained distributionof oil pools.