Perceptual hashing method for video content authentication with maximized robustness

EURASIP Journal on Image and Video Processing

Table 1 Robustness and sensitivity performances of the perceptual content preserving operations

Methods	Our method		LSH-core		Radon-based		TIRI + DCT
Methods	P_F (%)	P_T (%)	P_F (%)	P_T (%)	P_F (%)	P_T (%)	P_F (%)	P_T (%)
Rotation	2.17	100	2.28	100	2.51	100	2.64	100
Scaling	2.31	100	2.35	100	2.47	99.72	2.49	98.17
Translation	2.25	100	2.41	99.85	2.61	98.15	2.81	97.16
Gaussian noise	2.31	99.80	2.51	99.11	2.63	98.04	2.71	97.12
Salt and pepper	2.34	99.84	2.61	98.99	2.68	97.85	2.83	97.56
Average filtering	2.51	99.15	2.71	98.14	2.78	96.62	3.12	96.10
Intensity changing	2.43	98.84	2.54	98.17	2.58	97.84	2.65	97.27
Median filtering	2.48	98.57	2.98	97.16	3.08	95.10	3.47	94.02

Here shown in this table are the false positive rates and true positive rates for all four methods under various operations. The listed each operation has the averaged results for each method at each row. The higher the true positive rate, the better the robustness; likewise, the lower the false positive rate, the better the sensitivity the method has