Ongoing work is primarily in Fortran on multiple processors of the Cray Y-MP/C90. The goal is to optimize the hybrid algorithm as much as possible. This work is being driven partly by the DIMACS Challenge.
A Comparison of Parallel Algorithms for Connected Components, by John Greiner, from the Sixth Annual ACM Symposium on Parallel Algorithms and Architectures (SPAA), June 27-29, 1994. Paper, abstract, slides, and talk notes.
The technical report contains more details about the optimizations and code. The conference paper contains more recent information, including the hybrid algorithm. A book chapter version in preparation contains more explanation about how the algorithms work.
The NESL code used in John Greiner's papers includes versions of
Related Work: parallel vision and image processing articles in the SEL-HPC High Performance Computing Archive
Up to the Irregular Algorithms page or John Greiner's research.