Remote Sensing Image Fusion Algorithm Based on Two-Stream Fusion Network and Residual Channel Attention Mechanism

Table 3

Evaluation indicators of different remote sensing image fusion algorithms under GaoFen-2 satellite datasets.

Reference comparison
Methods	SCC	SSIM	Q	PSNR	ERGAS	SAM
Brovey	0.8914	0.7991	0.7452	42.0577	7.9374	0.2451
PCA	0.8364	0.7284	0.6446	42.2696	9.2124	0.2716
IHS	0.8831	0.7768	0.7236	43.5003	7.8865	0.1796
SFIM	0.8713	0.7132	0.7451	43.3909	9.0278	0.1865
MTF_GTP	0.9271	0.8021	0.7467	44.5458	6.2051	0.1932
CNMF	0.9281	0.8275	0.7717	44.9722	6.8851	0.1993
PNN	0.9471	0.8021	0.7631	46.4144	5.2051	0.1432
DPRNN	0.9489	0.8275	0.7747	46.9355	4.9311	0.1333
PanNet	0.9576	0.8619	0.8395	47.7915	4.8269	0.1364
Our method	0.9681	0.8656	0.8499	48.0889	4.6449	0.1216
Nonreference comparison
Methods					QNR
Brovey	0.0744		0.1968		0.7435
PCA	0.1047		0.1564		0.7553
IHS	0.1195		0.1165		0.7779
SFIM	0.0606		0.1485		0.7999
MTF_GLP	0.1069		0.1018		0.8022
CNMF	0.1061		0.0925		0.8112
PNN	0.1511		0.0319		0.8218
DPRNN	0.0900		0.0926		0.8275
PanNet	0.0734		0.0889		0.8442
Our method	0.0824		0.0532		0.8688