首页 | 本学科首页   官方微博 | 高级检索  
     

结合图像梯度和亮度的并行信任传播算法
引用本文:何俊学,李战明.结合图像梯度和亮度的并行信任传播算法[J].微电子学与计算机,2011,28(9):70-72,76.
作者姓名:何俊学  李战明
作者单位:兰州理工大学电信学院,甘肃兰州,730050
摘    要:对立体匹配问题建立马尔可夫随机场模型,使用并行的多尺度信任传播算法求解马尔可夫随机场的能量最小化问题.在传统串行算法基础上利用CUDA技术实现了并行计算,并结合图像的梯度和亮度信息计算能量函数的数据项,平滑项采用两个相邻像素视差的绝对差度量.以标准的Middlebury立体数据集做为输入,实验结果表明:算法具有很好的实时性能,运行时间远小于传统的串行算法,深度图结果较优.

关 键 词:计算机视觉  立体匹配  马尔可夫随机场  信任传播  并行计算

Parallel Belief Propagation Algorithm Conjugating Gradient and Intensity
HE Jun-xue,LI Zhan-ming.Parallel Belief Propagation Algorithm Conjugating Gradient and Intensity[J].Microelectronics & Computer,2011,28(9):70-72,76.
Authors:HE Jun-xue  LI Zhan-ming
Affiliation:HE Jun-xue,LI Zhan-ming(College of Electrical and Information Engineering,Lanzhou University of Technology,Lanzhou 730050,China)
Abstract:Stereo matching is critical technology in vision measurement.MRF models are established to do with stereo problem.A parallel multi-scale belief propagation algorithm is used for MRF energy minimization and generating disparity map.Parallel algorithm is implemented based on traditional sequential algorithm with CUDA technology.In energy function,data term is conjugated with Gradient and intensity of images,smooth term is measured with the absolute difference of disparities between two adjacent pixels.With st...
Keywords:computer vision  stereo matching  Markov random field  belief propagation  parallel computing  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号