图像融合效果评价方法研究

随着遥感技术的发展，遥感数据越来越丰富。通过遥感图像融合可以综合利用各种遥感数据的信息，提高遥感图像解译和自动分类的准确性。图像融合的方法很多，但缺乏对融合图像效果评价的方法来综合评价融合方法的效果。本文采用标准差、信息熵和清晰度等定量参数来评价融合效果并用自动分类的精度来验证。

Study on the Effect Evaluation of Image Fusion

As the remote sensing technology developing, there are many RS data can be received, and all kinds of RS data can be used to the image fusion; the image fusions could enhance precision of image translation and classification. There are many methods of image fusion, but it lacks an image fusion's evaluation system to evaluate their effect. The paper proposes to use standard deviation, entropy and definition to evaluate the effect of image fusion and to validate it by precision of classification.