基于TV-L1模型的多尺度遥感影像分解方法

TV—L^2模型是一种有效的遥感影像分解方法，但采用TV—L^2模型进行遥感影像分解会出现分解不彻底的情况。基于TV—L^1模型的多尺度遥感影像分解方法是一种将TV—L^1模型和多尽度方法结合起来对遥感影像进行分解的遥感影像分解方法。实验证明，基于TV—L^1模型的多尺度遥感影像分解方法继承了Tv—^L2模型利用全变分分解的优点，解决了TV—L^2模型分解不彻底的问题，并且通过尺度调节实现了多尺度分解。实验结果表明，该方法在遥感影像分解中取得了理想的效果，可以进一步地扩展应用。

A Multi-Scale Remote Sensing Image Decomposition MethodBased on TV-L1 Model

The TV-L^2 model has been proposed to serve as an effective remote sensing image decomposition method. Nevertheless, the image decomposition result of the TV-L^2 model is sometimesunsatisfactory. In this paper, a remote sensing image decomposition method combining multi-scale decomposition with the TV-L^1 model was proposed. Experimental results showed that our method not only retained the advantages of the TV-L^2 model, but also overcame the drawback of its decomposition result not clear. It can be further extended application, such as remote sensing image enhance or remote sensing image segmentation.