图像分解模型在医学图像增强中的应用

医学图像增强是医学图像处理中的重要环节。通过分析小波去噪和ROF模型的缺陷,先利用ROF分解模型将医学图像分解成为轮廓部分和细节及噪声部分,然后对轮廓部分进行保留,接着考虑到小波系数的非高斯性,对细节和噪声进行了小波去噪,并从中提取了图像的细节部分,最后将之前的轮廓部分与之后的细节部分进行叠加。实验结果表明,本文的算法具有较高的峰值信噪比和较高的边缘保持度。

Application of Image Decomposition Model to Medical Image Enhancement

Medical image enhancement is an important section during the medical image processing. Through analysising shortcomings of ROF（Rudin, Osber and Fatemi） model and Wavelet de-noising, first, we decomposed the image into three parts which are structure, details as well as noise, and kept the structure afterwards. Next, by considering on the non-Gaussian character of the wavelet coefficients, we applied Wavelet de-noising to the details as well as noise part. Then we obtained the details. Finally, we plused the structure and the details. According to the result,＇ the higher PSNR and higher EPI have witnessed the better performance of the algorithm.