Distribution strategies for very large 3D image deconvolution algorithms
作者：
Highlights：
•
• Running deconvolution algorithms with large images is not possible on a single node.
• Main issues are storage capacity and computation performance.
• Distributed implementation on a cluster allows memory load reduction on each node.
• The overall processing time is smaller on the cluster than on a single computer.
摘要
•Running deconvolution algorithms with large images is not possible on a single node.•Main issues are storage capacity and computation performance.•Distributed implementation on a cluster allows memory load reduction on each node.•The overall processing time is smaller on the cluster than on a single computer.
论文关键词：Image deconvolution,Distributed optimization algorithms,Large 3D data
论文评审过程：Received 4 December 2017, Revised 26 March 2018, Accepted 21 June 2018, Available online 26 June 2018, Version of Record 29 June 2018.
论文官网地址：https://doi.org/10.1016/j.image.2018.06.012
