卫星多源遥感图像融合研究

卫星多源遥感图像融合技术可以综合利用多源信息,从而获得对同一目标更全面、更清晰、更准确的认识。文中简介卫星多源遥感图像融合系统所包含的三个过程,即预处理、图像融合、综合评价及应用;详细分析像素层、特征层和决策层三个不同融合层次的融合原理和方法;归纳融合效果的主客观评价标准和熵、平均梯度等六种评价指标;最后从空间配准模型、建立统一的数学融合模型、关联的二义性、数据融合方法与融合系统实施存在的问题、有效的评价准则等方面讨论了卫星多源遥感图像融合目前存在的问题,并对今后的发展趋势作出展望。

Research on Satellite Multi-source Remote Sensing Image Fusion

The satellite multi-source remote sensing image fusion synthesizes multiple sensors images and information for the same region, thus the fused image can contain much more information and then it is more efficient, more accurate and clearer. Firstly this paper simply describes the three main procedures involved in the satellite multi-source remote sensing image fusion system, these are pre-processing, image fusion and comprehensive evaluation and application. Afterwards, this paper analyzes the fusion theories and methods of each fusion level, and gives out subjective and objective evaluation criteria and six kinds of methods for fusion images such as entropy and average gradient. Finally, The problems existed in the multi-source remote sensing image fusion are discussed from different aspects such as the model of space matching, the establish of uniform math fusion model,the problems exist in data fusion method and system, the efficient evaluation criterion etc. the development trend is also discussed.