小波基函数选择对SAR干涉图去噪的影响

The Influence on SAR Interferograms Noise Reduction due to the Selection of Wavelet Base Function

首先简要介绍Daubeachies、Biorthogonal、Symlets、Coiflets、Discrete Meyer和Reverse Biorthogonal六种小波函数的特性,然后以这些小波函数为基函数,采用相同的小波去噪算法对含噪声的模拟SAR干涉图和实际形变干涉图进行去噪实验。实验结果表明:在六种小波函数中,Biorthogonal小波的去噪效果最好,而且计算速度较快,是SAR干涉图去噪的最佳小波基函数。

The properties of six wavelet functions, such as Daubechies, Biorthogonal, Symlets, Coiflets, Discrete Meyer, Reverse Biorthogonal, are briefly introduced in this paper. Using these wavelet functions as base function, two denoising experiments have been performed through a simulated SAR interferogram contaminated by noise and a practical deformation interferogram, with the same noise reduction algorithm in wavelet domain. The experiment results show that Biorthogonal wavelet obviously outperforms the other five wavelet functions in term of the noise removal and its computing speed is relatively fast. Therefore, Biorthogonal wavelet is an optimal wavelet base function to noise reduction of the SAR interferogram.