Matching

Results: Table

Although our approach is different from the one in Siddiqi et al., our results are similar.

Similarity between database shapes and class prototypes

		Class Prototype
Instance
	->	245	2772	3436	3078	1274	1950	4840	8189	1631
	->	2515	898	4564	4347	2348	1864	5933	7638	3408
	->	3205	5721	3607	3982	2702	3921	3812	7005	3407
	->	3348	5346	1480	806	4191	5242	3896	6424	3161
	->	4868	5376	1537	1906	3890	5097	5402	7403	4137
	->	5279	5433	1789	526	3902	4790	3945	5921	3389
	->	5627	5429	1777	1545	3783	5801	4844	6095	4043
	->	774	2998	2724	4198	626	1465	4979	8862	2569
	->	1587	2814	2861	4749	652	1685	5633	8951	2336
	->	2779	5075	3737	4893	1567	258	4631	8057	1995
	->	6231	9451	4418	4332	4627	5759	519	4611	4145
	->	5007	8250	4860	4202	4218	4815	1998	5938	3522
	->	8668	12057	6201	5854	6295	7175	3186	2914	5720
	->	2406	4935	5385	3868	1772	2251	5320	6124	798
	->	1855	4296	3205	3222	1045	2528	4761	6054	877

Please compare this table to the one in Siddiqi, see [23]

Table2: Best Instance Matches

Prototype Best 2nd 3rd
out
of
range

out
of
range out
of
range
out
of
range

This table shows the best matches of the instances to each prototype.
A box marked 'out of range' means, the next best match owns a similarity value, which is greater than two times the value of the best match.

Return to matching: results
Continue with Formal Description