| 并行各向异性扩散算法与实时医学图像增强技术 |
| |
| 引用本文： | 秦安,孟晓林,冯前进,陈武凡.并行各向异性扩散算法与实时医学图像增强技术[J].医疗设备信息,2010(3):29-31,36. |
| |
| 作者姓名： | 秦安 孟晓林 冯前进 陈武凡 |
| |
| 作者单位： | 南方医科大学广东省医学图像处理重点实验室,广东广州510515 |
| |
| 基金项目： | 国家自然科学基金重点项目（30730036）;广东产学研项目（cgzhzdo714） |
| |
| 摘 要： | 各向异性扩散算法在去除图像噪声同时能保持重要的边缘和局部细节，因此在医学图像增强中得到广泛应用。但医学图像数据分辨率和灰度级都很高，求解各向异性扩散的偏微分方程时运算馈很大，在许多实时系统中应用该类算法存在照著迟滞。本文采用CUDA架构，利用像素的独啦性和偏微分方程求解的并发性，实现了并行各向异性扩散算法，在保证效果的同时显著降低了处理时间，满足了实时性的需求。
|
| 关 键 词： | 各向异性扩散方程 图像增强 CUDA 医学图像 |
Parallel Anisotropic Diffusion Algorithm and Real-time Medical Image Enhancement Technology |
| |
| QIN An,MENG Xiao-lin,FENG Qian-jin,CHEN Wu-fan.Parallel Anisotropic Diffusion Algorithm and Real-time Medical Image Enhancement Technology[J].Information of Medical Equipment,2010(3):29-31,36. |
| |
| Authors: | QIN An MENG Xiao-lin FENG Qian-jin CHEN Wu-fan |
| |
| Institution: | (Key Labs of Medical Image Processing in Guangdong, Southern Medical University, Guangzhou Guangdong 510515,China) |
| |
| Abstract: | Anisotropic diffusion can remore noiscs while preserving image boundaries, and has been widely used in medical image enhancement.But typical medical images have high resolution and grayscale, and the computation of the partial differential equation is huge and the time cost has blocked its applications in real-time medical imaging system. This paper implements anisotropie diffusion image enhancement under CUDA architecture. By taking use of the independence of image pixel and of partial differential equation, parallel computing in GPU can greatly reduce the computing time while keeping the good performance of image smoothing. |
| |
| Keywords: | parallel anisntrupie diffusion equation image enhancement CUDA medical image |
| 本文献已被 维普 等数据库收录！ |
