图像奇异性表征机理分析比较

Image Singularity Analysis and Comparison

图像奇异性包含的许多重要信息对于图像的进一步分析具有重要作用。对于图像中的不同奇异性通常需要采用不同的方法表示。小波变换和曲波变换作为稀疏表示中的重要方法,具有广泛的应用。分析了小波变换和曲波变换对于图像奇异性表证的不同效果。实验结果和理论分析均表明小波变换对于图像中的点奇异性具有很好效果,但对于线奇异性表示则不够稀疏,曲波则可以高效地表示图像边缘的曲线奇异性。

Singularity in images carry massive essential information,which is crucial to further analysis.We need to adopt different methods according to the elements in image singularity.Wavelet transform and curvelet transform are two important methods for sparse representation and are widely used in numerous fields.In this paper,wavelet transform and curvelet transform are used to analyze the effects of different singularity in images.The experimental results and theoretical analysis both demonstrate that wavelet is good at noises in images but cannot represent the edges effectively.On the contrary,curvelet is very suitable for the curve edge.