#clccl An OpenCL implementation of Connected Components Labeling based on "Connected Component Labeling in CUDA" from Onrej Stava, Bedrich Benes (2010) with speedups, bugfixes et al.
Instructions:
place clcommons in the current directory
Benchmark:
python -m bench -i $ITERATIONS <image>
unit tests:
python -m unit_tests <image>