基于Curvelet变换的自适应阈值图像去噪方法

Adaptive threshold image denoising methed based on curvelet transform

与小波变换相比,Curvelet变换能更好地表达图像的边缘和细节,因此更适合做图像处理.提出了一种基于第二代Curvelet变换的自适应阈值图像去噪方法,采用不同的阈值自适应地对不同尺度和方向的Curvelet系数进行阈值处理.实验结果表明,提出的方法在去除噪声的同时,能更好地保留图像的细节.去噪后的图像有更高的峰值信噪比和更好的视觉效果.

The curvelet transform represents edges and details of image better than wavelet transform, and is therefore well - suited for image processing. A adaptive threshold image denoising method based on the second curvelet transform is proposed. The curvelet transform coefficients in different scales and directions are filtered with adaptive threshold. Experiments show that the method proposed yields denoised image with higher quality recovery of edges. It is capable of achieving the higher peak signal - noise - ratio and giving better visual quality.