A framework for measuring sharpness in natural images captured by digital cameras based on reference image and local areas

EURASIP Journal on Image and Video Processing

Table 3 The content-specific validation performance values for Dataset I. Boldface indicates the best performer.

Content	Metric	ROCC		LCC		RMSE		RMSE (subjective)
		Training	Validation	Training	Validation	Training	Validation
Content 1	Proposed metric	0.857	0.863	0.879	0.884	7.953	10.700	21.651
	Wang [18]	0.783	0.692	0.814	0.625	9.711	12.818
Content 2	Proposed metric	0.801	0.901	0.806	0.901	8.461	8.022	25.647
	Wang [18]	0.667	0.379	0.665	-0.484	10.677	8.907
Content 3	Proposed metric	0.847	0.478	0.851	0.725	8.514	8.060	16.110
	Wang [18]	0.712	0.676	0.730	0.782	11.084	6.177
Content 4	Proposed metric	0.892	0.626	0.900	0.664	7.601	11.056	20.231
	Wang [18]	0.835	0.209	0.829	0.362	9.746	12.450
Content 5	Proposed metric	0.866	0.912	0.875	0.913	8.338	8.723	20.045
	Wang [18]	0.832	0.797	0.833	0.770	9.548	14.158