基于双边滤波算子的医学脊椎去噪

发布时间：2018-06-26 21:06

本文选题：三维模型 + 双边滤波　；参考：《光子学报》2017年07期

【摘要】：为了有效滤除医学脊椎模型的噪声点,同时更好地保持模型细节,提出了一种基于双边滤波算子的医学脊椎去噪模型.采用双边滤波在多尺度条件下进行脊椎三维模型轮廓线的提取,设计改进自适应扩散系数,以更好的优化控制整个扩散过程.根据图像的离散特征,建立相应的离散迭代方程,使迭代过程离散化,并设计迭代停止准则,当去噪平滑后的图像模型与噪声相关性最小时停止迭代.与经典的向异性扩散模型方法实验结果相比,本方法在解决去噪方面达到了很好的滤波效果,同时也较好地保持了医学图像的边缘细节特征,大大优于传统滤波算法.
[Abstract]:In order to effectively filter the noise points of the medical spinal model and keep the details of the model better, a medical spine denoising model based on the bilateral filtering operator is proposed. Two-sided filter is used to extract the contour of the three-dimensional model of spine under the condition of multi-scale, and the adaptive diffusion coefficient is improved to optimize the whole diffusion process. According to the discrete characteristics of the image, the corresponding discrete iterative equations are established to discretize the iterative process, and the iterative stopping criterion is designed. When the image model after denoising and smoothing has the least correlation with noise, the iteration is stopped. Compared with the experimental results of the classical anisotropic diffusion model method, this method achieves a good filtering effect in solving the denoising problem, and at the same time, it also maintains the edge detail feature of the medical image, which is much better than the traditional filtering algorithm.
【作者单位】：西北工业大学计算机学院;西北工业大学软件与微电子学院;西安电子科技大学计算机学院;西北工业大学管理学院;法士特集团咸阳精密机械分公司;
【基金】：国家自然科学基金(No.61172147)资助~~
【分类号】：R318;TP391.41

【相似文献】