首页> 外文期刊>Computers & mathematics with applications >Lattice Boltzmann method for filtering and contour detection of the natural images
【24h】

Lattice Boltzmann method for filtering and contour detection of the natural images

机译:格子Boltzmann方法对自然图像进行滤波和轮廓检测

获取原文
获取原文并翻译 | 示例

摘要

In this paper, the lattice Boltzmann method (LBM) is extended to study the filtering and contour detection of natural images, and a new lattice Boltzmann model is proposed for more complicated image processing model, like the Ambrosio and Tortorelli (A-T) model that contains two coupled nonlinear partial differential equations. The numerical results of image filtering and contour detection show that the noises in the image can be removed greatly, and simultaneously, important contours of the image are protected well. To improve the computational efficiency, we implement the developed lattice Boltzmann model on Graphic Processing Unit (GPU), and find that, compared to the CPU based algorithm, the GPU based LBM can gain more than 25 × speedup, which is very important in the further lattice Boltzmann study of large-scale image processing problems. And finally, these numerical results also show that the LBM is a feasible and efficient approach for filtering and contour detection of the natural images.
机译:本文扩展了格子Boltzmann方法(LBM)来研究自然图像的滤波和轮廓检测,并针对复杂的图像处理模型(如Ambrosio和Tortorelli(AT)模型)提出了新的格子Boltzmann模型两个耦合的非线性偏微分方程。图像滤波和轮廓检测的数值结果表明,可以大大消除图像中的噪声,同时,很好地保护了图像的重要轮廓。为了提高计算效率,我们在图形处理单元(GPU)上实现了已开发的格子Boltzmann模型,发现与基于CPU的算法相比,基于GPU的LBM可以实现25倍以上的加速,这对于提高计算效率非常重要。进一步的格子Boltzmann研究了大型图像处理问题。最后,这些数值结果还表明,LBM是一种对自然图像进行滤波和轮廓检测的可行而有效的方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号