轮廓波及曲波和小波变换用于显微图像消噪的比较*****☆

背景：小波变换只能反映信号的零维奇异性,无法最优表示图像中的线奇异;而且小波变换只存在3个方向,这些都显著影响了它在图像处理领域的应用效果。针对小波变换的缺点,多尺度几何分析理论正在逐步发展,轮廓波变换和曲波变换就是其中的典型代表。目的：定性、定量地比较轮廓波、曲波和小波变换在图像消噪处理中的效果。方法：在简要介绍3种变换基本原理的基础上,比较它们在图像消噪领域的应用,以均方误差和峰值信噪比作为定量指标评价消噪效果,并将其应用于显微镜图像的消噪处理。结果与结论：综合定量评价指标和人眼视觉感受,曲波变换的消噪结果最佳,轮廓波变换效果次之,小波变换效果则不够理想。

Comparison of microscopy image denoising effects based on contourlet, curvelet and wavelet transform

BACKGROUND: Wavelets in two-dimension are good at isolating the discontinuities at edge points, but not the smoothness along the contours. In addition, separable wavelets only capture limited directional information, which weaken their application effects on image processing. Image multiscale geometric analysis theory is developed gradually to overcome the shortcomings of wavelets mentioned above.OBJECTIVE: To compare the microscopy image denoising effects qualitatively and quantitatively based on contourlet, curvelet and wavelet transforms. METHODS: Based on the brief descriptions of contourlet, curvelet and wavelet transform, performance analysis and comparison were done depending on image denoising with qualitative and quantitative indices computed, e.g., mean square error and peak signal-to-noise ratio.RESULTS AND CONCLUSION: Experimental results demonstrate that for the test Lena images and microscopy images, curvelet transform achieves the best result, while wavelet transform result is poor.