从第2代曲波变换为基础，提出了一种利用尺度相关性进行自适应阈值收缩的图像去噪方法。通过实验比较小波与第2代曲波图像去噪，采用峰值信噪比对去噪结果进行了客观评价，实验结果表明，本文提出的图像去噪方法明显优于小波图像去噪方法，并且相对于单纯的阈值收缩去噪也有很大提高。

In this paper, an adaptive thresholding image denoising method with scale correlation based on the second generation Curvelet transform is proposed. Experiments compare image denoising based on Wavelet and the second generation Curvelet transform. PSNR is used to evaluate the results, and the results show that the method based on the second generation Curvelet transform achieves better denoising than wavelet and simple thresholding methods.