Abstract: According to the noise characteristic and application requirement of Computed Tomography（CT） slice image, the denoising algorithm based on Lipschitz exponent and Stationary Wavelet Transform（SWT） is proposed. This algorithm uses the improved median filter for filtering the impulse noise, then based on the relationship of Lipschitz exponent of step edge and wavelet coefficient and under the premise of protecting the image edge details, the threshold denoising method based on SWT for filtering Gaussian noise is adopted. The experimental results show that whether the visual effects or the perfomance of the minimum Mean Square Error（MSE） and the Peak of Signal-to-Noise Ratio（PSNR）, and protecting the image edge details of this algorithm have great improvement.

摘要： 针对CT切片图像噪声特点以及应用需求，提出了基于Lipschitz指数和平稳小波的CT图像去噪算法。利用改进的中值滤波器滤除图像的脉冲噪声，然后根据图像阶梯边缘的Lipschitz指数与小波系数之间的关系，在更好地保护图像边缘细节的前提下，利用平稳小波变换的阈值去噪方法滤除高斯噪声。实验结果表明，该方法无论是在视觉效果上，还是在最小均方差意义和信噪比增益上，以及保护图像边缘细节上都有很大提高。