A fast source-oriented image clustering method for digital forensics

EURASIP Journal on Image and Video Processing

Table 5 Comparison of clustering algorithms on \(\mathcal {D}_{5}\)

Algorithms	1024×1024					512×512					512×256
	\(\mathcal {P}\)	\(\mathcal {R}\)	\(\mathcal {F}\)	\(\mathcal {A}\)	\(\mathcal {N}\)	\(\mathcal {P}\)	\(\mathcal {R}\)	\(\mathcal {F}\)	\(\mathcal {A}\)	\(\mathcal {N}\)	\(\mathcal {P}\)	\(\mathcal {R}\)	\(\mathcal {F}\)	\(\mathcal {A}\)	\(\mathcal {N}\)
SC [31]	0.03	0.64	0.06	0.00	0.03	0.17	0.94	0.29	0.04	0.09	0.36	0.80	0.50	0.15	0.24
HC [34]	0.46	0.65	0.54	0.22	0.50	0.81	0.31	0.45	0.65	1.72	0.78	0.15	0.26	0.44	3.35
SNNC [43]	0.47	0.27	0.34	0.24	1.08	0.44	0.20	0.28	0.21	1.38	0.35	0.14	0.20	0.10	1.55
NCUT [45]	0.89	0.13	0.23	0.58	4.68	0.73	0.18	0.29	0.51	2.77	0.58	0.19	0.29	0.36	2.01
MRF [49]	0.60	0.54	0.57	0.29	0.70	0.55	0.55	0.55	0.24	0.62	0.43	0.57	0.49	0.16	0.49
Proposed	0.88	0.79	0.83	0.79	0.74	0.74	0.76	0.75	0.58	0.61	0.63	0.65	0.64	0.40	0.58