【摘要】 为了提高SAR影像的解译水平,避免通常基于小波变换的融合方法造成的SAR影像信息损失,本文提出一种基于a’trous小波与广义HIS变换的SAR与多光谱影像融合方法,在将多光谱影像转换到HIS空间后,应用a’trous小波对I分量进行分解,通过加法的形式将多光谱影像的高频分量信息与SAR影像信息集成,并根据解译的需要,通过改变阈值来控制对多光谱影像信息的集成幅度。实验选取一组TM多光谱影像与ERS-2SAR影像进行融合研究,并将融合结果与另一小波融合方法融合结果进行视觉比较与统计分析。结果表明,另一小波融合方法的融合结果与本文方法融合结果阈值τ=1时的结果接近,而本文方法却可以根据不同的应用需要,在完整保留了SAR影像信息的基础上,通过调节多光谱影像信息的注入程度,为获取更能满足解译需要的SAR融合影像提供更多选择,拥有更好的鲁棒性。更多还原

【Abstract】 A fusion method of SAR and multi-spectral Images based on a’trous Wavelet and Generalized HIS Transform with threshold is proposed in this paper in order to improve SAR images’ interpretation and to avoid information loss when using wavelet transform to fuse SAR image and Multispectral images in some applications.I component was decomposed by a’trous wavelet after multispectral images were transformed to HIS space and high frequency information of I was added to SAR image.And integration degree of multispectral image information is controlled by variation of threshold according to needs of interpretation.A data set of TM images and ERS-2 SAR image is used to evaluate the proposed method.The fusion results are evaluated visually and statistically and compared to another method based on discrete wavelet and HIS transform.The evaluation shows that another method’s result is close to the fused image of this paper’s method when threshold τ=1,however,the method proposed in this paper could be used according to different needs,uses threshold to control the extent of multispectral images information and preserves information of SAR image completely,provides more choices to produce better fused SAR images to use,besides it is more robust.更多还原