SAR Image Matching Based on Local Feature Detection and Description Using Convolutional Neural Network

Table 2

Performance results of first dataset.

Study case	Metric	Before	CC	SIFT	ORB	Proposed	[12]
Subset	Shift	−4, −6	0, −1	0, 0	0, 0	0, 0	0, 0
	MSE	103.24	102.06	39.67	53.19	67.89	90.37
	PSNR	13.979	14.021	16.073	15.436	14.906	14.285
	CC	0.059	0.278	0.621	0.572	0.552	0.344
Translation	Shift	8, −1	3, 0	0, 0	0, 0	0, 0	0, 0
	MSE	103.97	104.16	70.39	87.15	81.65	101.15
	PSNR	13.981	13.976	14.827	14.364	14.505	14.041
	CC	−0.079	−0.062	0.429	0.333	0.376	0.031
Scaling	Shift	0, 0	0, 0	0, 0	0, 0	0, 0	0, 0
	MSE	4.09	81.27	58.88	92.51	84.17	94.62
	PSNR	21.001	14.515	15.215	14.234	14.439	14.185
	CC	−0.093	0.901	0.938	0.618	0.853	0.428
Rotation	Shift	10, −2	10, 8	0, 0	0, 0	0, 0	0, 0
	MSE	103.91	103.59	88.32	91.68	103.98	103.62
	PSNR	13.982	13.988	14.335	14.254	13.981	13.988
	CC	−0.142	−0.153	0.475	0.444	−0.018	−0.119