Parallelization of the Hoshen-Kopelman Algorithm Using a Finite State Machine

Abstract

In applications such as landscape ecology, computer mod eling is used to assess habitat fragmentation and its ecological implications. Maps (two-dimensional grids) of habitat clusters or patches are analyzed to determine the number, location, and sizes of clusters. Recently, improved sequential and parallel implementations of the Hoshen- Kopelman cluster identification algorithm have been designed. These implementations use a finite state ma chine to reduce redundant integer comparisons during the cluster identification process. The sequential implementa tion for large maps performs cluster identification by par titioning the map along row boundaries and merging the results of the partitions. The parallel implementation on a 32-processor Thinking Machines CM-5 provides an effi cient mechanism for performing cluster identification in parallel. Although the sequential implementation achieved promising speed improvements ranging from 1.39 to 2.00 over an existing Hoshen-Kopelman implementation, the parallel implementation achieved a minimum speedup of 5.41 over the improved sequential implementation, exe cuted on a Sun SPARCstation 10.

Finite state machine-based optimization of data parallel regular domain problems applied in low-level image processing

“…As in our case, results indicate that the fsm approach requires very little runtime overhead. For ad hoc optimization of specific algorithms and applications, fsm definitions have been applied successfully as well [5], [18].…”

Multi-GPU-based Swendsen–Wang multi-cluster algorithm with reduced data traffic

“…These approaches all execute the cluster labeling on each CPU using the Hoshen-Kopelman algorithm, and then perform the cluster labeling between each CPU. The approach to cluster labeling between each CPU differs in each method: there is a master-slave method [17], a method that prepares a global label array [23], and a relaxation method that exchanges boundary information between neighbors until all labels are unchanged [24,25]. As a refinement of this relaxation method, Bauernfeind et al [26] proposed a technique that reduced the data traffic.…”

Performance of a Finite-State Machine Implementation of Iterative Cluster Labeling on Desktop and Mobile Computing Platforms

“…The implementation of HK presented in [10] utilizes a finite-state machine to achieve improved performance and is the basis for the original research presented in this paper. However, one fundamental difference between this FSM implementation (that we will simply refer to as FSM-HK) and that presented in Section 3 is the use of the nearest-four neighborhood rule.…”

“…The HK algorithm has previously been implemented using a finite-state machine (FSM) to improve upon its performance, but that implementation is limited to a neighborhood rule in which only the four cardinal neighbors are considered to be connected to a point in the map [10]. Due to artifacts introduced by the process of rasterizing a landscape map, it is often desirable to consider a point in a map to be connected to its nearest eight neighbors [11].…”