Fig. 1. Framework of proposed remote sensing image mosaic algorithm
Fig. 2. Diagram of bidirectional mutual selection matching
Fig. 3. Weighted fusion process
Fig. 4. Response curve of human visual perception
Fig. 5. Plant growth function curve
Fig. 6. Function curve of weighting factor ω1
Fig. 7. Weighted fusion process of improved S-model
Fig. 8. Preprocessing of remote sensing image. (a) Original image;(b) mean filtering; (c) guided filtering
Fig. 9. Comparison of feature point matching strategies. (a) One-way matching; (b) bidirectional mutual selection matching; (c) partial enlargement of Fig. 9(a); (d) partial enlargement of Fig. 9(b)
Fig. 10. Comparison of matching accuracy for images
Fig. 11. Comparison of three fusion strategies for images of industrial_area. (a)(b) Original images; (c) S-type fusion; (d) direct average fusion; (e) gradual integration
Fig. 12. Comparison of three fusion strategies for images of intersection. (a)(b) Original images; (c) S-type fusion; (d) direct average fusion; (e) gradual integration
Fig. 13. Comparison of three fusion strategies for images of harbor. (a)(b) Original images; (c) S-type fusion; (d) direct average fusion; (e) gradual integration
Fig. 14. Comparison of three fusion strategies for remote sensing images obtained from Gaofen-1 satellite with resolution of 2 m/8 m. (a)(b) Original images; (c) S-type fusion; (d) direct average fusion; (e) gradual integration
Fig. 15. Comparison of three fusion strategies for remote sensing images obtained from Gaofen-1 satellite with resolution of 16 m. (a)(b) Original images; (c) S-type fusion; (d) direct average fusion; (e) gradual integration
Fig. 16. Comparison of image information entropy data
Matching strategy | Matching point type | Number of matching points |
---|
Group 1 | Group 2 | Group 3 |
---|
One-way matching | Rough matching points | 7078 | 8016 | 12220 | Final matching points | 2269 | 3093 | 4347 | Bidirectional mutualselection matching | Rough matching points from left to right | 5887 | 7618 | 10395 | Matching points from left to right | 2260 | 3092 | 4324 | Rough matching points from right to left | 7078 | 8016 | 12220 | Matching points from right to left | 2269 | 3093 | 4341 | Final matching points | 2246 | 3090 | 4309 |
|
Table 1. Comparison of experimental data for matching points of images
Fusion strategy | Average gradient |
---|
Group 1 | Group 2 | Group 3 |
---|
S-type fusion algorithm | 8.3934 | 7.0391 | 7.4159 | Direct averagefusion algorithm | 4.0435 | 3.8064 | 3.0115 | Progressive weightedfusion algorithm | 4.0454 | 3.8079 | 3.0136 |
|
Table 2. Comparison of image average gradient data