摘要
针对传统数字图像处理中匹配方法虽然精确度高,但计算量大、时间长等特点,将基于MPI(Message Passing Interface)的集群并行处理引入到图像灰度匹配中,对待匹配图像采用数据分割处理,而将传统的图像灰度匹配算法进行并行化改进,结合并行处理对图像灰度匹配进行并行实现.实验结果表明:并行化处理能显著地缩短灰度匹配时间,达到较高的加速比和效率,对进一步研究基于集群系统下的并行图像处理有一定的指导意义.
Gray scale matching plays an important role in digital image processing. Although high precision may be achieved using the traditional gray scale matching methods, the results are attained at the cost of large time-consumption and heavy computational load. To improve the matching efficiency, the message passing interface (MPI) is introduced in this paper. The cluster parallel processing is adopted in image gray scale matching, in which the parallel computing techniques are utilized to improve the efficiency of the traditional image gray scale matching algorithm by means as of fusion of data division processing. The proposed scheme is put to the test with results showing that parallel processing is significantly time-saving in gray scale matching, and sufficiently simplified with its computational complexity. This paper is expected to bring some technical insights into further investigation of cluster system based parallel image processing.
出处
《宁波大学学报(理工版)》
CAS
2009年第1期74-77,共4页
Journal of Ningbo University:Natural Science and Engineering Edition
基金
宁波大学科研项目(200582XK200583)
关键词
MPI
并行图像处理
灰度匹配
集群
MPI
parallel image processing
gray scale matching
cluster