Abstract: Due to the image details of the Contourlet transform have vibration at the singularity points which would lead to Gibbs-like phenomena.A method for image denoising based on the Contourlet transform with improved Laplacian Pyramid（LP） is proposed.The selection of threshold value considers the difference of noise content in different scales，and also makes the adjustment in the different directions to the threshold value.The experimental results indicate that this method can get better visual effect and PSNR value compared with others.

Key words: image denoising, Contourlet transform, Laplacian pyramid, adaptive threshold

摘要： 针对Contourlet分解的细节图像在奇异点附近产生振荡，在去噪过程中会产生伪吉布斯现象，提出一种改进的拉普拉斯金字塔实现基于Contourlet变换的图像去噪算法。阈值的选取不仅考虑不同尺度中噪声含量的不同，而且在不同方向上对阈值进行了调整。实验结果表明，利用该文去噪方法进行去噪比其他方法得到更好的视觉效果和更高的PSNR值。

关键词: 图像去噪, Contourlet变换, 拉普拉斯金字塔, 自适应阈值