返回
MapReduce-based fast fuzzy c-means algorithm for large-scale underwater image segmentation
DOI:10.1016/j.future.2016.03.004.png)
摘要
En 中文
The research on underwater image segmentation has to deal with the rapid increasing volume of images and videos. To handle this issue, parallel computing paradigms, such as the MapReduce framework has been proven as a viable solution. Therefore, we propose a MapReduce-based fast fuzzy c-means algorithm (MRFFCM) to paralyze the segmentation of the images. In our work, we use a two-layer distribution model to group the large-scale images and adopt an iterative MapReduce process to parallelize the FFCM algorithm. A combinational segmentation way is used to improve algorithm's efficiency. To evaluate the performance of our algorithm, we develop a small Hadoop cluster to test the MRFFCM algorithm. The experiment results demonstrate that our proposed method is effective and efficient on large-scale images. When compared to the traditional non-parallel methods, our algorithm can be expected to provide a more efficient segmentation on images with at least 13% improvement. Meanwhile, with the growth of cluster size, further improvement of the algorithm performance was also achieved. Consequently, such scalability can enable our proposed method to be used effectively in oceanic research, such as in underwater data processing systems. (C) 2016 Elsevier B.V. All rights reserved.
Keyword:
Fast fuzzy c-means algorithm
Image segmentation
MapReduce
Two-layer distribution model
AI总结
对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
F
IF:
6.1
论文数:
6.8K
被引数:
2.3W
机构
引用论文
BBR: An open-source standard workflow based on biophysical crop parameters for automatic Botrytis cinerea assessment in vineyards
SoftwareX
IF0

