中文摘要：

针对传统水体指数在水体提取中存在的问题，提出了一种综合水体指数CWI；并将其应用于ETM＋和SPOT5遥感图像上水体的提取实验。实验结果表明，在经过CWI运算产生的综合水体指数图像上，水体与非水体的灰度值有明显的差异，能较好地区分水体与居民地、阴影等背景地物，无需多次试验确定阈值，水体提取效果好，且能适用于多种遥感图像。

英文摘要：

ACWI（Comprehensive Water Index） was put forward to resolve the problems of traditional water index in water extraction. It was applied to water body extraction on ETM＋ and SPOT5 imagery. The experi- ments results showed that there is obvious difference between water pixel intensity and non-water pixel intensity on the CWI image. Water body and background features such as residential areas and shadows can be well distinguished and threshold can be automatically determined. CWI has better results and is appropriate for various remote sensing imagery.