GF-1 Image Fusion Based on Regression Kriging

Shutao Wang; Wang Kang; Deming Kong; Tiezhu Wang; Ruixiang Li

doi:10.3788/LOP202259.0828005

Journals >Laser & Optoelectronics Progress >Volume 59 >Issue 8 >Page 0828005 > Article

Laser & Optoelectronics Progress
Vol. 59, Issue 8, 0828005 (2022)

GF-1 Image Fusion Based on Regression Kriging

Shutao Wang¹, Wang Kang^1、*, Deming Kong¹, Tiezhu Wang¹, and Ruixiang Li²

Author Affiliations

¹Key Laboratory of Measurement and Measurement Technology and Instruments, Electrical Engineering College,Yanshan University, Qinhuangdao , Hebei 066004, China

²School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo , Henan 454003, China

show less

DOI: 10.3788/LOP202259.0828005 Cite this Article Set citation alerts

Shutao Wang, Wang Kang, Deming Kong, Tiezhu Wang, Ruixiang Li. GF-1 Image Fusion Based on Regression Kriging[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0828005 Copy Citation Text

show less

Fig. 1. Flow chart of proposed method

Download full size

Fig. 2. Experiment GF-1 image and location of selected experimental area

Download full size

Fig. 3. Region 1 fusion results of each algorithm. (a) Source data; (b) PAN image; (c) IGS; (d) PCA; (e) WaveLets; (f) DenseNet; (g) ATPRK; (h) proposed algorithm

Download full size

Fig. 4. Region 2 fusion results of each algorithm. (a) Source data; (b) PAN image; (c) IGS; (d) PCA; (e) WaveLets; (f) DenseNet;(g) ATPRK; (h) proposed algorithm

Download full size

Fig. 5. Region 3 fusion results of each algorithm. (a) Source data; (b) PAN image; (c) IGS; (d) PCA; (e) WaveLets; (f) DenseNet;(g) ATPRK; (h) proposed algorithm

Download full size

Fig. 6. Local amplification of fusion results of each algorithm.(a) Source data; (b) PAN image; (c) IGS; (d) PCA; (e) WaveLets;(f) DenseNet; (g) ATPRK;(h) proposed algorithm

Download full size

Index	Band	Ideal	IGS	PCA	WaveLets	DenseNet	ATPRK	Proposed algorithm
RMSE	R	0	0.0184	0.0358	0.0174	0.0148	0.0121	0.0092
	G	0	0.0197	0.0424	0.018	0.016	0.0157	0.0125
	B	0	0.0193	0.0503	0.0179	0.0118	0.0174	0.0137
	NIR	0	0.0195	0.006	0.1148	0.0119	0.0113	0.0124
	Mean	0	0.0192	0.0336	0.0420	0.0136	0.0141	0.0118
SSIM	R	1	0.9353	0.1474	0.9573	0.9307	0.9808	0.9842
	G	1	0.9450	0.1777	0.9663	0.9014	0.9776	0.9787
	B	1	0.9540	0.1616	0.9727	0.9405	0.9797	0.9822
	NIR	1	0.9521	0.9918	0.827	0.8984	0.981	0.9805
	Mean	1	0.9466	0.3696	0.9308	0.9178	0.9798	0.9817
UIQI	R	1	0.4903	0.5269	0.8706	0.9345	0.913	0.9442
	G	1	0.5042	0.5415	0.8992	0.9163	0.9008	0.9304
	B	1	0.6137	0.5376	0.9316	0.9352	0.92	0.9513
	NIR	1	0.5455	0.9925	0.8119	0.9571	0.9372	0.9322
	Mean	1	0.5384	0.6496	0.8783	0.9358	0.9178	0.9394
ERGAS		0	9.9768	21.330	24.541	6.4678	7.7983	6.1853
SAM		0	1.2179	4.9176	13.380	1.2254	1.433	1.2354

Table 1. Results of image quality evaluation in experimental region 1

Index	Band	Ideal	IGS	PCA	WaveLets	DenseNet	ATPRK	Proposed algorithm
RMSE	R	0	0.0183	0.034	0.0179	0.0096	0.0114	0.0095
	G	0	0.0195	0.0412	0.0184	0.0136	0.0157	0.0125
	B	0	0.0192	0.0451	0.0182	0.0101	0.0159	0.0132
	NIR	0	0.0195	0.0119	0.094	0.0113	0.0176	0.0125
	Mean	0	0.0191	0.0330	0.0371	0.0112	0.0152	0.0118
SSIM	R	1	0.9584	0.2779	0.9803	0.9507	0.9722	0.9954
	G	1	0.9642	0.308	0.9851	0.9638	0.9677	0.9941
	B	1	0.9684	0.351	0.9877	0.9461	0.9706	0.9949
	NIR	1	0.9637	0.945	0.7077	0.9848	0.9651	0.9948
	Mean	1	0.9637	0.4705	0.9152	0.9614	0.9689	0.9949
UIQI	R	1	0.8903	0.4461	0.8687	0.9265	0.923	0.9428
	G	1	0.9177	0.4734	0.9071	0.9634	0.9057	0.9442
	B	1	0.9404	0.5184	0.9308	0.9368	0.928	0.9515
	NIR	1	0.9233	0.9653	0.8148	0.9769	0.893	0.9637
	Mean	1	0.9179	0.6008	0.8803	0.9509	0.9124	0.9503
ERGAS		0	9.6995	19.065	21.819	7.1079	7.7372	6.0844
SAM		0	1.2404	3.5079	9.8272	1.3506	1.6204	1.1741

Table 2. Results of image quality evaluation in experimental region 2

Index	Bands	Ideal	IGS	PCA	WaveLets	DenseNet	ATPRK	Proposed algorithm
RMSE	R	0	0.0126	0.0071	0.0123	0.0074	0.0025	0.0026
	G	0	0.0133	0.0064	0.0125	0.0066	0.0046	0.0046
	B	0	0.0132	0.0097	0.0125	0.0088	0.0043	0.0048
	NIR	0	0.0182	0.0409	0.0819	0.0125	0.013	0.0101
	Mean	0	0.0143	0.0160	0.0298	0.0088	0.0061	0.0055
SSIM	R	1	0.8369	0.5515	0.9143	0.9814	0.9898	0.9964
	G	1	0.8874	0.8144	0.9551	0.9747	0.9847	0.9945
	B	1	0.8833	0.5889	0.9508	0.9921	0.9866	0.9954
	NIR	1	0.9368	0.2124	0.8426	0.9814	0.9833	0.9942
	Mean	1	0.8861	0.5418	0.9157	0.9824	0.9861	0.9952
UIQI	R	1	0.6344	0.7855	0.6318	0.9025	0.9254	0.9191
	G	1	0.6247	0.8933	0.7367	0.9256	0.9051	0.8983
	B	1	0.6576	0.8037	0.7656	0.8959	0.9176	0.922
	NIR	1	0.7637	0.6	0.6695	0.9387	0.936	0.9611
	Mean	1	0.6701	0.7706	0.7009	0.9157	0.9210	0.9248
ERGAS		0	10.232	9.8129	34.957	3.7847	3.8249	3.5724
SAM		0	1.7334	4.5052	27.873	1.1054	1.0947	0.8931

Table 3. Results of image quality evaluation in experimental region 3

Shutao Wang, Wang Kang, Deming Kong, Tiezhu Wang, Ruixiang Li. GF-1 Image Fusion Based on Regression Kriging[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0828005

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information