Zhang-Suen-Thinning-Algorithm

This is the implementation of the Zhang-Suen thinning algorithm. It's implemented with two different ways. A serial and a parallel version. The parallel version is implemented with the MPI standard and more specifically with the MPICH implementation.

   

Processes	Execution Time (s)	Acceleration (Sp=Ts/Tp)	Efficiency (Sp/p)	Communication Cost (s)
1	0.1272912	1	1	-
2	0.1172022	1.08	0.54	0.0053226
4	0.085999	1.48	0.37	0.0154506
8	0.2005016	0.634	0.079	0.1286832
16	0.4238256	0.3	0.018	0.3372468

   

Processes	Execution Time (s)	Acceleration (Sp=Ts/Tp)	Efficiency (Sp/p)	Communication Cost (s)
1	0.0727708	1.03	1.03	-
2	0.064187	1.13	0.565	0.002727
4	0.0805224	0.903	0.225	0.0066016
8	0.2064952	0.352	0.044	0.1831272
16	0.4732442	0.153	0.009	0.4098266

   

Processes	Execution Time (s)	Acceleration (Sp=Ts/Tp)	Efficiency (Sp/p)	Communication Cost (s)
1	2.1640772	0.98	0.98	-
2	1.144749	1.89	0.945	0.0409786
4	0.7248446	2.98	0.748	0.1296998
8	0.6203674	3.488	0.436	0.2193812
16	0.7055856	3.067	0.191	0.3879998

Tested on Ubuntu 20.04.

Requirements

• Ubuntu 18.04 or newer.
• GCC compiler.
• MPICH.

References

The algorithm is based on this paper.